You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

648 lines
29 KiB

8 years ago
8 years ago
Account domain blocks (#2381) * Add <ostatus:conversation /> tag to Atom input/output Only uses ref attribute (not href) because href would be the alternate link that's always included also. Creates new conversation for every non-reply status. Carries over conversation for every reply. Keeps remote URIs verbatim, generates local URIs on the fly like the rest of them. * Conversation muting - prevents notifications that reference a conversation (including replies, favourites, reblogs) from being created. API endpoints /api/v1/statuses/:id/mute and /api/v1/statuses/:id/unmute Currently no way to tell when a status/conversation is muted, so the web UI only has a "disable notifications" button, doesn't work as a toggle * Display "Dismiss notifications" on all statuses in notifications column, not just own * Add "muted" as a boolean attribute on statuses JSON For now always false on contained reblogs, since it's only relevant for statuses returned from the notifications endpoint, which are not nested Remove "Disable notifications" from detailed status view, since it's only relevant in the notifications column * Up max class length * Remove pending test for conversation mute * Add tests, clean up * Rename to "mute conversation" and "unmute conversation" * Raise validation error when trying to mute/unmute status without conversation * Adding account domain blocks that filter notifications and public timelines * Add tests for domain blocks in notifications, public timelines Filter reblogs of blocked domains from home * Add API for listing and creating account domain blocks * API for creating/deleting domain blocks, tests for Status#ancestors and Status#descendants, filter domain blocks from them * Filter domains in streaming API * Update account_domain_block_spec.rb
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Account domain blocks (#2381) * Add <ostatus:conversation /> tag to Atom input/output Only uses ref attribute (not href) because href would be the alternate link that's always included also. Creates new conversation for every non-reply status. Carries over conversation for every reply. Keeps remote URIs verbatim, generates local URIs on the fly like the rest of them. * Conversation muting - prevents notifications that reference a conversation (including replies, favourites, reblogs) from being created. API endpoints /api/v1/statuses/:id/mute and /api/v1/statuses/:id/unmute Currently no way to tell when a status/conversation is muted, so the web UI only has a "disable notifications" button, doesn't work as a toggle * Display "Dismiss notifications" on all statuses in notifications column, not just own * Add "muted" as a boolean attribute on statuses JSON For now always false on contained reblogs, since it's only relevant for statuses returned from the notifications endpoint, which are not nested Remove "Disable notifications" from detailed status view, since it's only relevant in the notifications column * Up max class length * Remove pending test for conversation mute * Add tests, clean up * Rename to "mute conversation" and "unmute conversation" * Raise validation error when trying to mute/unmute status without conversation * Adding account domain blocks that filter notifications and public timelines * Add tests for domain blocks in notifications, public timelines Filter reblogs of blocked domains from home * Add API for listing and creating account domain blocks * API for creating/deleting domain blocks, tests for Status#ancestors and Status#descendants, filter domain blocks from them * Filter domains in streaming API * Update account_domain_block_spec.rb
7 years ago
8 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Non-Serial ("Snowflake") IDs (#4801) * Use non-serial IDs This change makes a number of nontrivial tweaks to the data model in Mastodon: * All IDs are now 8 byte integers (rather than mixed 4- and 8-byte) * IDs are now assigned as: * Top 6 bytes: millisecond-resolution time from epoch * Bottom 2 bytes: serial (within the millisecond) sequence number * See /lib/tasks/db.rake's `define_timestamp_id` for details, but note that the purpose of these changes is to make it difficult to determine the number of objects in a table from the ID of any object. * The Redis sorted set used for the feed will have values used to look up toots, rather than scores. This is almost always the same as the existing behavior, except in the case of boosted toots. This change was made because Redis stores scores as double-precision floats, which cannot store the new ID format exactly. Note that this doesn't cause problems with sorting/pagination, because ZREVRANGEBYSCORE sorts lexicographically when scores are tied. (This will still cause sorting issues when the ID gains a new significant digit, but that's extraordinarily uncommon.) Note a couple of tradeoffs have been made in this commit: * lib/tasks/db.rake is used to enforce many/most column constraints, because this commit seems likely to take a while to bring upstream. Enforcing a post-migrate hook is an easier way to maintain the code in the interim. * Boosted toots will appear in the timeline as many times as they have been boosted. This is a tradeoff due to the way the feed is saved in Redis at the moment, but will be handled by a future commit. This would effectively close Mastodon's #1059, as it is a snowflake-like system of generating IDs. However, given how involved the changes were simply within Mastodon, it may have unexpected interactions with some clients, if they store IDs as doubles (or as 4-byte integers). This was a problem that Twitter ran into with their "snowflake" transition, particularly in JavaScript clients that treated IDs as JS integers, rather than strings. It therefore would be useful to test these changes at least in the web interface and popular clients before pushing them to all users. * Fix JavaScript interface with long IDs Somewhat predictably, the JS interface handled IDs as numbers, which in JS are IEEE double-precision floats. This loses some precision when working with numbers as large as those generated by the new ID scheme, so we instead handle them here as strings. This is relatively simple, and doesn't appear to have caused any problems, but should definitely be tested more thoroughly than the built-in tests. Several days of use appear to support this working properly. BREAKING CHANGE: The major(!) change here is that IDs are now returned as strings by the REST endpoints, rather than as integers. In practice, relatively few changes were required to make the existing JS UI work with this change, but it will likely hit API clients pretty hard: it's an entirely different type to consume. (The one API client I tested, Tusky, handles this with no problems, however.) Twitter ran into this issue when introducing Snowflake IDs, and decided to instead introduce an `id_str` field in JSON responses. I have opted to *not* do that, and instead force all IDs to 64-bit integers represented by strings in one go. (I believe Twitter exacerbated their problem by rolling out the changes three times: once for statuses, once for DMs, and once for user IDs, as well as by leaving an integer ID value in JSON. As they said, "If you’re using the `id` field with JSON in a Javascript-related language, there is a very high likelihood that the integers will be silently munged by Javascript interpreters. In most cases, this will result in behavior such as being unable to load or delete a specific direct message, because the ID you're sending to the API is different than the actual identifier associated with the message." [1]) However, given that this is a significant change for API users, alternatives or a transition time may be appropriate. 1: https://blog.twitter.com/developer/en_us/a/2011/direct-messages-going-snowflake-on-sep-30-2011.html * Restructure feed pushes/unpushes This was necessary because the previous behavior used Redis zset scores to identify statuses, but those are IEEE double-precision floats, so we can't actually use them to identify all 64-bit IDs. However, it leaves the code in a much better state for refactoring reblog handling / coalescing. Feed-management code has been consolidated in FeedManager, including: * BatchedRemoveStatusService no longer directly manipulates feed zsets * RemoveStatusService no longer directly manipulates feed zsets * PrecomputeFeedService has moved its logic to FeedManager#populate_feed (PrecomputeFeedService largely made lots of calls to FeedManager, but didn't follow the normal adding-to-feed process.) This has the effect of unifying all of the feed push/unpush logic in FeedManager, making it much more tractable to update it in the future. Due to some additional checks that must be made during, for example, batch status removals, some Redis pipelining has been removed. It does not appear that this should cause significantly increased load, but if necessary, some optimizations are possible in batch cases. These were omitted in the pursuit of simplicity, but a batch_push and batch_unpush would be possible in the future. Tests were added to verify that pushes happen under expected conditions, and to verify reblog behavior (both on pushing and unpushing). In the case of unpushing, this includes testing behavior that currently leads to confusion such as Mastodon's #2817, but this codifies that the behavior is currently expected. * Rubocop fixes I could swear I made these changes already, but I must have lost them somewhere along the line. * Address review comments This addresses the first two comments from review of this feature: https://github.com/tootsuite/mastodon/pull/4801#discussion_r139336735 https://github.com/tootsuite/mastodon/pull/4801#discussion_r139336931 This adds an optional argument to FeedManager#key, the subtype of feed key to generate. It also tests to ensure that FeedManager's settings are such that reblogs won't be tracked forever. * Hardcode IdToBigints migration columns This addresses a comment during review: https://github.com/tootsuite/mastodon/pull/4801#discussion_r139337452 This means we'll need to make sure that all _id columns going forward are bigints, but that should happen automatically in most cases. * Additional fixes for stringified IDs in JSON These should be the last two. These were identified using eslint to try to identify any plain casts to JavaScript numbers. (Some such casts are legitimate, but these were not.) Adding the following to .eslintrc.yml will identify casts to numbers: ~~~ no-restricted-syntax: - warn - selector: UnaryExpression[operator='+'] > :not(Literal) message: Avoid the use of unary + - selector: CallExpression[callee.name='Number'] message: Casting with Number() may coerce string IDs to numbers ~~~ The remaining three casts appear legitimate: two casts to array indices, one in a server to turn an environment variable into a number. * Only implement timestamp IDs for Status IDs Per discussion in #4801, this is only being merged in for Status IDs at this point. We do this in a migration, as there is no longer use for a post-migration hook. We keep the initialization of the timestamp_id function as a Rake task, as it is also needed after db:schema:load (as db/schema.rb doesn't store Postgres functions). * Change internal streaming payloads to stringified IDs as well This is equivalent to 591a9af356faf2d5c7e66e3ec715502796c875cd from #5019, with an extra change for the addition to FeedManager#unpush. * Ensure we have a status_id_seq sequence Apparently this is not a given when specifying a custom ID function, so now we ensure it gets created. This uses the generic version of this function to more easily support adding additional tables with timestamp IDs in the future, although it would be possible to cut this down to a less generic version if necessary. It is only run during db:schema:load or the relevant migration, so the overhead is extraordinarily minimal. * Transition reblogs to new Redis format This provides a one-way migration to transition old Redis reblog entries into the new format, with a separate tracking entry for reblogs. It is not invertible because doing so could (if timestamp IDs are used) require a database query for each status in each users' feed, which is likely to be a significant toll on major instances. * Address review comments from @akihikodaki No functional changes. * Additional review changes * Heredoc cleanup * Run db:schema:load hooks for test in development This matches the behavior in Rails' ActiveRecord::Tasks::DatabaseTasks.each_current_configuration, which would otherwise break `rake db:setup` in development. It also moves some functionality out to a library, which will be a good place to put additional related functionality in the near future.
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
8 years ago
8 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Web Push Notifications (#3243) * feat: Register push subscription * feat: Notify when mentioned * feat: Boost, favourite, reply, follow, follow request * feat: Notification interaction * feat: Handle change of public key * feat: Unsubscribe if things go wrong * feat: Do not send normal notifications if push is enabled * feat: Focus client if open * refactor: Move push logic to WebPushSubscription * feat: Better title and body * feat: Localize messages * chore: Fix lint errors * feat: Settings * refactor: Lazy load * fix: Check if push settings exist * feat: Device-based preferences * refactor: Simplify logic * refactor: Pull request feedback * refactor: Pull request feedback * refactor: Create /api/web/push_subscriptions endpoint * feat: Spec PushSubscriptionController * refactor: WebPushSubscription => Web::PushSubscription * feat: Spec Web::PushSubscription * feat: Display first media attachment * feat: Support direction * fix: Stuff broken while rebasing * refactor: Integration with session activations * refactor: Cleanup * refactor: Simplify implementation * feat: Set VAPID keys via environment * chore: Comments * fix: Crash when no alerts * fix: Set VAPID keys in testing environment * fix: Follow link * feat: Notification actions * fix: Delete previous subscription * chore: Temporary logs * refactor: Move migration to a later date * fix: Fetch the correct session activation and misc bugs * refactor: Move migration to a later date * fix: Remove follow request (no notifications) * feat: Send administrator contact to push service * feat: Set time-to-live * fix: Do not show sensitive images * fix: Reducer crash in error handling * feat: Add badge * chore: Fix lint error * fix: Checkbox label overlap * fix: Check for payload support * fix: Rename action "type" (crash in latest Chrome) * feat: Action to expand notification * fix: Lint errors * fix: Unescape notification body * fix: Do not allow boosting if the status is hidden * feat: Add VAPID keys to the production sample environment * fix: Strip HTML tags from status * refactor: Better error messages * refactor: Handle browser not implementing the VAPID protocol (Samsung Internet) * fix: Error when target_status is nil * fix: Handle lack of image * fix: Delete reference to invalid subscriptions * feat: Better error handling * fix: Unescape HTML characters after tags are striped * refactor: Simpify code * fix: Modify to work with #4091 * Sort strings alphabetically * i18n: Updated Polish translation it annoys me that it's not fully localized :P * refactor: Use current_session in PushSubscriptionController * fix: Rebase mistake * fix: Set cacheName to mastodon * refactor: Pull request feedback * refactor: Remove logging statements * chore(yarn): Fix conflicts with master * chore(yarn): Copy latest from master * chore(yarn): Readd offline-plugin * refactor: Use save! and update! * refactor: Send notifications async * fix: Allow retry when push fails * fix: Save track for failed pushes * fix: Minify sw.js * fix: Remove account_id from fabricator
7 years ago
Web Push Notifications (#3243) * feat: Register push subscription * feat: Notify when mentioned * feat: Boost, favourite, reply, follow, follow request * feat: Notification interaction * feat: Handle change of public key * feat: Unsubscribe if things go wrong * feat: Do not send normal notifications if push is enabled * feat: Focus client if open * refactor: Move push logic to WebPushSubscription * feat: Better title and body * feat: Localize messages * chore: Fix lint errors * feat: Settings * refactor: Lazy load * fix: Check if push settings exist * feat: Device-based preferences * refactor: Simplify logic * refactor: Pull request feedback * refactor: Pull request feedback * refactor: Create /api/web/push_subscriptions endpoint * feat: Spec PushSubscriptionController * refactor: WebPushSubscription => Web::PushSubscription * feat: Spec Web::PushSubscription * feat: Display first media attachment * feat: Support direction * fix: Stuff broken while rebasing * refactor: Integration with session activations * refactor: Cleanup * refactor: Simplify implementation * feat: Set VAPID keys via environment * chore: Comments * fix: Crash when no alerts * fix: Set VAPID keys in testing environment * fix: Follow link * feat: Notification actions * fix: Delete previous subscription * chore: Temporary logs * refactor: Move migration to a later date * fix: Fetch the correct session activation and misc bugs * refactor: Move migration to a later date * fix: Remove follow request (no notifications) * feat: Send administrator contact to push service * feat: Set time-to-live * fix: Do not show sensitive images * fix: Reducer crash in error handling * feat: Add badge * chore: Fix lint error * fix: Checkbox label overlap * fix: Check for payload support * fix: Rename action "type" (crash in latest Chrome) * feat: Action to expand notification * fix: Lint errors * fix: Unescape notification body * fix: Do not allow boosting if the status is hidden * feat: Add VAPID keys to the production sample environment * fix: Strip HTML tags from status * refactor: Better error messages * refactor: Handle browser not implementing the VAPID protocol (Samsung Internet) * fix: Error when target_status is nil * fix: Handle lack of image * fix: Delete reference to invalid subscriptions * feat: Better error handling * fix: Unescape HTML characters after tags are striped * refactor: Simpify code * fix: Modify to work with #4091 * Sort strings alphabetically * i18n: Updated Polish translation it annoys me that it's not fully localized :P * refactor: Use current_session in PushSubscriptionController * fix: Rebase mistake * fix: Set cacheName to mastodon * refactor: Pull request feedback * refactor: Remove logging statements * chore(yarn): Fix conflicts with master * chore(yarn): Copy latest from master * chore(yarn): Readd offline-plugin * refactor: Use save! and update! * refactor: Send notifications async * fix: Allow retry when push fails * fix: Save track for failed pushes * fix: Minify sw.js * fix: Remove account_id from fabricator
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
Make IdsToBigints (mostly!) non-blocking (#5088) * Make IdsToBigints (mostly!) non-blocking This pulls in GitLab's MigrationHelpers, which include code to make column changes in ways that Postgres can do without locking. In general, this involves creating a new column, adding an index and any foreign keys as appropriate, adding a trigger to keep it populated alongside the old column, and then progressively copying data over to the new column, before removing the old column and replacing it with the new one. A few changes to GitLab's MigrationHelpers were necessary: * Some changes were made to remove dependencies on other GitLab code. * We explicitly wait for index creation before forging ahead on column replacements. * We use different temporary column names, to avoid running into index name length limits. * We rename the generated indices back to what they "should" be after replacing columns. * We rename the generated foreign keys to use the new column names when we had to create them. (This allows the migration to be rolled back without incident.) # Big Scary Warning There are two things here that may trip up large instances: 1. The change for tables' "id" columns is not concurrent. In particular, the stream_entries table may be big, and does not concurrently migrate its id column. (On the other hand, x_id type columns are all concurrent.) 2. This migration will take a long time to run, *but it should not lock tables during that time* (with the exception of the "id" columns as described above). That means this should probably be run in `screen` or some other session that can be run for a long time. Notably, the migration will take *longer* than it would without these changes, but the website will still be responsive during that time. These changes were tested on a relatively large statuses table (256k entries), and the service remained responsive during the migration. Migrations both forward and backward were tested. * Rubocop fixes * MigrationHelpers: Support ID columns in some cases This doesn't work in cases where the ID column is referred to as a foreign key by another table. * MigrationHelpers: support foreign keys for ID cols Note that this does not yet support foreign keys on non-primary-key columns, but Mastodon also doesn't yet have any that we've needed to migrate. This means we can perform fully "concurrent" migrations to change ID column types, and the IdsToBigints migration can happen with effectively no downtime. (A few operations require a transaction, such as renaming columns or deleting them, but these transactions should not block for noticeable amounts of time.) The algorithm for generating foreign key names has changed with this, and therefore all of those changed in schema.rb. * Provide status, allow for interruptions The MigrationHelpers now allow restarting the rename of a column if it was interrupted, by removing the old "new column" and re-starting the process. Along with this, they now provide status updates on the changes which are happening, as well as indications about when the changes can be safely interrupted (when there are at least 10 seconds estimated to be left before copying data is complete). The IdsToBigints migration now also sorts the columns it migrates by size, starting with the largest tables. This should provide administrators a worst-case scenario estimate for the length of migrations: each successive change will get faster, giving admins a chance to abort early on if they need to run the migration later. The idea is that this does not force them to try to time interruptions between smaller migrations. * Fix column sorting in IdsToBigints Not a significant change, but it impacts the order of columns in the database and db/schema.rb. * Actually pause before IdsToBigints
7 years ago
  1. # This file is auto-generated from the current state of the database. Instead
  2. # of editing this file, please use the migrations feature of Active Record to
  3. # incrementally modify your database, and then regenerate this schema definition.
  4. #
  5. # Note that this schema.rb definition is the authoritative source for your
  6. # database schema. If you need to create the application database on another
  7. # system, you should be using db:schema:load, not running all the migrations
  8. # from scratch. The latter is a flawed and unsustainable approach (the more migrations
  9. # you'll amass, the slower it'll run and the greater likelihood for issues).
  10. #
  11. # It's strongly recommended that you check this file into your version control system.
  12. ActiveRecord::Schema.define(version: 2018_07_11_152640) do
  13. # These are extensions that must be enabled in order to support this database
  14. enable_extension "plpgsql"
  15. create_table "account_domain_blocks", force: :cascade do |t|
  16. t.string "domain"
  17. t.datetime "created_at", null: false
  18. t.datetime "updated_at", null: false
  19. t.bigint "account_id"
  20. t.index ["account_id", "domain"], name: "index_account_domain_blocks_on_account_id_and_domain", unique: true
  21. end
  22. create_table "account_moderation_notes", force: :cascade do |t|
  23. t.text "content", null: false
  24. t.bigint "account_id", null: false
  25. t.bigint "target_account_id", null: false
  26. t.datetime "created_at", null: false
  27. t.datetime "updated_at", null: false
  28. t.index ["account_id"], name: "index_account_moderation_notes_on_account_id"
  29. t.index ["target_account_id"], name: "index_account_moderation_notes_on_target_account_id"
  30. end
  31. create_table "accounts", force: :cascade do |t|
  32. t.string "username", default: "", null: false
  33. t.string "domain"
  34. t.string "secret", default: "", null: false
  35. t.text "private_key"
  36. t.text "public_key", default: "", null: false
  37. t.string "remote_url", default: "", null: false
  38. t.string "salmon_url", default: "", null: false
  39. t.string "hub_url", default: "", null: false
  40. t.datetime "created_at", null: false
  41. t.datetime "updated_at", null: false
  42. t.text "note", default: "", null: false
  43. t.string "display_name", default: "", null: false
  44. t.string "uri", default: "", null: false
  45. t.string "url"
  46. t.string "avatar_file_name"
  47. t.string "avatar_content_type"
  48. t.integer "avatar_file_size"
  49. t.datetime "avatar_updated_at"
  50. t.string "header_file_name"
  51. t.string "header_content_type"
  52. t.integer "header_file_size"
  53. t.datetime "header_updated_at"
  54. t.string "avatar_remote_url"
  55. t.datetime "subscription_expires_at"
  56. t.boolean "silenced", default: false, null: false
  57. t.boolean "suspended", default: false, null: false
  58. t.boolean "locked", default: false, null: false
  59. t.string "header_remote_url", default: "", null: false
  60. t.integer "statuses_count", default: 0, null: false
  61. t.integer "followers_count", default: 0, null: false
  62. t.integer "following_count", default: 0, null: false
  63. t.datetime "last_webfingered_at"
  64. t.string "inbox_url", default: "", null: false
  65. t.string "outbox_url", default: "", null: false
  66. t.string "shared_inbox_url", default: "", null: false
  67. t.string "followers_url", default: "", null: false
  68. t.integer "protocol", default: 0, null: false
  69. t.boolean "memorial", default: false, null: false
  70. t.bigint "moved_to_account_id"
  71. t.string "featured_collection_url"
  72. t.jsonb "fields"
  73. t.string "actor_type"
  74. t.index "(((setweight(to_tsvector('simple'::regconfig, (display_name)::text), 'A'::\"char\") || setweight(to_tsvector('simple'::regconfig, (username)::text), 'B'::\"char\")) || setweight(to_tsvector('simple'::regconfig, (COALESCE(domain, ''::character varying))::text), 'C'::\"char\")))", name: "search_index", using: :gin
  75. t.index "lower((username)::text), lower((domain)::text)", name: "index_accounts_on_username_and_domain_lower", unique: true
  76. t.index ["uri"], name: "index_accounts_on_uri"
  77. t.index ["url"], name: "index_accounts_on_url"
  78. end
  79. create_table "admin_action_logs", force: :cascade do |t|
  80. t.bigint "account_id"
  81. t.string "action", default: "", null: false
  82. t.string "target_type"
  83. t.bigint "target_id"
  84. t.text "recorded_changes", default: "", null: false
  85. t.datetime "created_at", null: false
  86. t.datetime "updated_at", null: false
  87. t.index ["account_id"], name: "index_admin_action_logs_on_account_id"
  88. t.index ["target_type", "target_id"], name: "index_admin_action_logs_on_target_type_and_target_id"
  89. end
  90. create_table "backups", force: :cascade do |t|
  91. t.bigint "user_id"
  92. t.string "dump_file_name"
  93. t.string "dump_content_type"
  94. t.integer "dump_file_size"
  95. t.datetime "dump_updated_at"
  96. t.boolean "processed", default: false, null: false
  97. t.datetime "created_at", null: false
  98. t.datetime "updated_at", null: false
  99. end
  100. create_table "blocks", force: :cascade do |t|
  101. t.datetime "created_at", null: false
  102. t.datetime "updated_at", null: false
  103. t.bigint "account_id", null: false
  104. t.bigint "target_account_id", null: false
  105. t.string "uri"
  106. t.index ["account_id", "target_account_id"], name: "index_blocks_on_account_id_and_target_account_id", unique: true
  107. end
  108. create_table "bookmarks", force: :cascade do |t|
  109. t.bigint "account_id", null: false
  110. t.bigint "status_id", null: false
  111. t.datetime "created_at", null: false
  112. t.datetime "updated_at", null: false
  113. t.index ["account_id", "status_id"], name: "index_bookmarks_on_account_id_and_status_id", unique: true
  114. t.index ["account_id"], name: "index_bookmarks_on_account_id"
  115. t.index ["status_id"], name: "index_bookmarks_on_status_id"
  116. end
  117. create_table "conversation_mutes", force: :cascade do |t|
  118. t.bigint "conversation_id", null: false
  119. t.bigint "account_id", null: false
  120. t.index ["account_id", "conversation_id"], name: "index_conversation_mutes_on_account_id_and_conversation_id", unique: true
  121. end
  122. create_table "conversations", force: :cascade do |t|
  123. t.string "uri"
  124. t.datetime "created_at", null: false
  125. t.datetime "updated_at", null: false
  126. t.index ["uri"], name: "index_conversations_on_uri", unique: true
  127. end
  128. create_table "custom_emojis", force: :cascade do |t|
  129. t.string "shortcode", default: "", null: false
  130. t.string "domain"
  131. t.string "image_file_name"
  132. t.string "image_content_type"
  133. t.integer "image_file_size"
  134. t.datetime "image_updated_at"
  135. t.datetime "created_at", null: false
  136. t.datetime "updated_at", null: false
  137. t.boolean "disabled", default: false, null: false
  138. t.string "uri"
  139. t.string "image_remote_url"
  140. t.boolean "visible_in_picker", default: true, null: false
  141. t.index ["shortcode", "domain"], name: "index_custom_emojis_on_shortcode_and_domain", unique: true
  142. end
  143. create_table "custom_filters", force: :cascade do |t|
  144. t.bigint "account_id"
  145. t.datetime "expires_at"
  146. t.text "phrase", default: "", null: false
  147. t.string "context", default: [], null: false, array: true
  148. t.boolean "irreversible", default: false, null: false
  149. t.boolean "whole_word", default: true, null: false
  150. t.datetime "created_at", null: false
  151. t.datetime "updated_at", null: false
  152. t.index ["account_id"], name: "index_custom_filters_on_account_id"
  153. end
  154. create_table "domain_blocks", force: :cascade do |t|
  155. t.string "domain", default: "", null: false
  156. t.datetime "created_at", null: false
  157. t.datetime "updated_at", null: false
  158. t.integer "severity", default: 0
  159. t.boolean "reject_media", default: false, null: false
  160. t.index ["domain"], name: "index_domain_blocks_on_domain", unique: true
  161. end
  162. create_table "email_domain_blocks", force: :cascade do |t|
  163. t.string "domain", default: "", null: false
  164. t.datetime "created_at", null: false
  165. t.datetime "updated_at", null: false
  166. t.index ["domain"], name: "index_email_domain_blocks_on_domain", unique: true
  167. end
  168. create_table "favourites", force: :cascade do |t|
  169. t.datetime "created_at", null: false
  170. t.datetime "updated_at", null: false
  171. t.bigint "account_id", null: false
  172. t.bigint "status_id", null: false
  173. t.index ["account_id", "id"], name: "index_favourites_on_account_id_and_id"
  174. t.index ["account_id", "status_id"], name: "index_favourites_on_account_id_and_status_id", unique: true
  175. t.index ["status_id"], name: "index_favourites_on_status_id"
  176. end
  177. create_table "follow_requests", force: :cascade do |t|
  178. t.datetime "created_at", null: false
  179. t.datetime "updated_at", null: false
  180. t.bigint "account_id", null: false
  181. t.bigint "target_account_id", null: false
  182. t.boolean "show_reblogs", default: true, null: false
  183. t.string "uri"
  184. t.index ["account_id", "target_account_id"], name: "index_follow_requests_on_account_id_and_target_account_id", unique: true
  185. end
  186. create_table "follows", force: :cascade do |t|
  187. t.datetime "created_at", null: false
  188. t.datetime "updated_at", null: false
  189. t.bigint "account_id", null: false
  190. t.bigint "target_account_id", null: false
  191. t.boolean "show_reblogs", default: true, null: false
  192. t.string "uri"
  193. t.index ["account_id", "target_account_id"], name: "index_follows_on_account_id_and_target_account_id", unique: true
  194. end
  195. create_table "identities", id: :serial, force: :cascade do |t|
  196. t.integer "user_id"
  197. t.string "provider", default: "", null: false
  198. t.string "uid", default: "", null: false
  199. t.datetime "created_at", null: false
  200. t.datetime "updated_at", null: false
  201. t.index ["user_id"], name: "index_identities_on_user_id"
  202. end
  203. create_table "imports", force: :cascade do |t|
  204. t.integer "type", null: false
  205. t.boolean "approved", default: false, null: false
  206. t.datetime "created_at", null: false
  207. t.datetime "updated_at", null: false
  208. t.string "data_file_name"
  209. t.string "data_content_type"
  210. t.integer "data_file_size"
  211. t.datetime "data_updated_at"
  212. t.bigint "account_id", null: false
  213. end
  214. create_table "invites", force: :cascade do |t|
  215. t.bigint "user_id", null: false
  216. t.string "code", default: "", null: false
  217. t.datetime "expires_at"
  218. t.integer "max_uses"
  219. t.integer "uses", default: 0, null: false
  220. t.datetime "created_at", null: false
  221. t.datetime "updated_at", null: false
  222. t.boolean "autofollow", default: false, null: false
  223. t.index ["code"], name: "index_invites_on_code", unique: true
  224. t.index ["user_id"], name: "index_invites_on_user_id"
  225. end
  226. create_table "list_accounts", force: :cascade do |t|
  227. t.bigint "list_id", null: false
  228. t.bigint "account_id", null: false
  229. t.bigint "follow_id", null: false
  230. t.index ["account_id", "list_id"], name: "index_list_accounts_on_account_id_and_list_id", unique: true
  231. t.index ["follow_id"], name: "index_list_accounts_on_follow_id"
  232. t.index ["list_id", "account_id"], name: "index_list_accounts_on_list_id_and_account_id"
  233. end
  234. create_table "lists", force: :cascade do |t|
  235. t.bigint "account_id", null: false
  236. t.string "title", default: "", null: false
  237. t.datetime "created_at", null: false
  238. t.datetime "updated_at", null: false
  239. t.index ["account_id"], name: "index_lists_on_account_id"
  240. end
  241. create_table "media_attachments", force: :cascade do |t|
  242. t.bigint "status_id"
  243. t.string "file_file_name"
  244. t.string "file_content_type"
  245. t.integer "file_file_size"
  246. t.datetime "file_updated_at"
  247. t.string "remote_url", default: "", null: false
  248. t.datetime "created_at", null: false
  249. t.datetime "updated_at", null: false
  250. t.string "shortcode"
  251. t.integer "type", default: 0, null: false
  252. t.json "file_meta"
  253. t.bigint "account_id"
  254. t.text "description"
  255. t.index ["account_id"], name: "index_media_attachments_on_account_id"
  256. t.index ["shortcode"], name: "index_media_attachments_on_shortcode", unique: true
  257. t.index ["status_id"], name: "index_media_attachments_on_status_id"
  258. end
  259. create_table "mentions", force: :cascade do |t|
  260. t.bigint "status_id"
  261. t.datetime "created_at", null: false
  262. t.datetime "updated_at", null: false
  263. t.bigint "account_id"
  264. t.index ["account_id", "status_id"], name: "index_mentions_on_account_id_and_status_id", unique: true
  265. t.index ["status_id"], name: "index_mentions_on_status_id"
  266. end
  267. create_table "mutes", force: :cascade do |t|
  268. t.datetime "created_at", null: false
  269. t.datetime "updated_at", null: false
  270. t.boolean "hide_notifications", default: true, null: false
  271. t.bigint "account_id", null: false
  272. t.bigint "target_account_id", null: false
  273. t.index ["account_id", "target_account_id"], name: "index_mutes_on_account_id_and_target_account_id", unique: true
  274. end
  275. create_table "notifications", force: :cascade do |t|
  276. t.bigint "activity_id", null: false
  277. t.string "activity_type", null: false
  278. t.datetime "created_at", null: false
  279. t.datetime "updated_at", null: false
  280. t.bigint "account_id", null: false
  281. t.bigint "from_account_id", null: false
  282. t.index ["account_id", "activity_id", "activity_type"], name: "account_activity", unique: true
  283. t.index ["account_id", "id"], name: "index_notifications_on_account_id_and_id", order: { id: :desc }
  284. t.index ["activity_id", "activity_type"], name: "index_notifications_on_activity_id_and_activity_type"
  285. end
  286. create_table "oauth_access_grants", force: :cascade do |t|
  287. t.string "token", null: false
  288. t.integer "expires_in", null: false
  289. t.text "redirect_uri", null: false
  290. t.datetime "created_at", null: false
  291. t.datetime "revoked_at"
  292. t.string "scopes"
  293. t.bigint "application_id", null: false
  294. t.bigint "resource_owner_id", null: false
  295. t.index ["token"], name: "index_oauth_access_grants_on_token", unique: true
  296. end
  297. create_table "oauth_access_tokens", force: :cascade do |t|
  298. t.string "token", null: false
  299. t.string "refresh_token"
  300. t.integer "expires_in"
  301. t.datetime "revoked_at"
  302. t.datetime "created_at", null: false
  303. t.string "scopes"
  304. t.bigint "application_id"
  305. t.bigint "resource_owner_id"
  306. t.index ["refresh_token"], name: "index_oauth_access_tokens_on_refresh_token", unique: true
  307. t.index ["resource_owner_id"], name: "index_oauth_access_tokens_on_resource_owner_id"
  308. t.index ["token"], name: "index_oauth_access_tokens_on_token", unique: true
  309. end
  310. create_table "oauth_applications", force: :cascade do |t|
  311. t.string "name", null: false
  312. t.string "uid", null: false
  313. t.string "secret", null: false
  314. t.text "redirect_uri", null: false
  315. t.string "scopes", default: "", null: false
  316. t.datetime "created_at"
  317. t.datetime "updated_at"
  318. t.boolean "superapp", default: false, null: false
  319. t.string "website"
  320. t.string "owner_type"
  321. t.bigint "owner_id"
  322. t.index ["owner_id", "owner_type"], name: "index_oauth_applications_on_owner_id_and_owner_type"
  323. t.index ["uid"], name: "index_oauth_applications_on_uid", unique: true
  324. end
  325. create_table "preview_cards", force: :cascade do |t|
  326. t.string "url", default: "", null: false
  327. t.string "title", default: "", null: false
  328. t.string "description", default: "", null: false
  329. t.string "image_file_name"
  330. t.string "image_content_type"
  331. t.integer "image_file_size"
  332. t.datetime "image_updated_at"
  333. t.integer "type", default: 0, null: false
  334. t.text "html", default: "", null: false
  335. t.string "author_name", default: "", null: false
  336. t.string "author_url", default: "", null: false
  337. t.string "provider_name", default: "", null: false
  338. t.string "provider_url", default: "", null: false
  339. t.integer "width", default: 0, null: false
  340. t.integer "height", default: 0, null: false
  341. t.datetime "created_at", null: false
  342. t.datetime "updated_at", null: false
  343. t.string "embed_url", default: "", null: false
  344. t.index ["url"], name: "index_preview_cards_on_url", unique: true
  345. end
  346. create_table "preview_cards_statuses", id: false, force: :cascade do |t|
  347. t.bigint "preview_card_id", null: false
  348. t.bigint "status_id", null: false
  349. t.index ["status_id", "preview_card_id"], name: "index_preview_cards_statuses_on_status_id_and_preview_card_id"
  350. end
  351. create_table "relays", force: :cascade do |t|
  352. t.string "inbox_url", default: "", null: false
  353. t.boolean "enabled", default: false, null: false
  354. t.string "follow_activity_id"
  355. t.datetime "created_at", null: false
  356. t.datetime "updated_at", null: false
  357. t.index ["enabled"], name: "index_relays_on_enabled"
  358. end
  359. create_table "report_notes", force: :cascade do |t|
  360. t.text "content", null: false
  361. t.bigint "report_id", null: false
  362. t.bigint "account_id", null: false
  363. t.datetime "created_at", null: false
  364. t.datetime "updated_at", null: false
  365. t.index ["account_id"], name: "index_report_notes_on_account_id"
  366. t.index ["report_id"], name: "index_report_notes_on_report_id"
  367. end
  368. create_table "reports", force: :cascade do |t|
  369. t.bigint "status_ids", default: [], null: false, array: true
  370. t.text "comment", default: "", null: false
  371. t.boolean "action_taken", default: false, null: false
  372. t.datetime "created_at", null: false
  373. t.datetime "updated_at", null: false
  374. t.bigint "account_id", null: false
  375. t.bigint "action_taken_by_account_id"
  376. t.bigint "target_account_id", null: false
  377. t.bigint "assigned_account_id"
  378. t.index ["account_id"], name: "index_reports_on_account_id"
  379. t.index ["target_account_id"], name: "index_reports_on_target_account_id"
  380. end
  381. create_table "session_activations", force: :cascade do |t|
  382. t.string "session_id", null: false
  383. t.datetime "created_at", null: false
  384. t.datetime "updated_at", null: false
  385. t.string "user_agent", default: "", null: false
  386. t.inet "ip"
  387. t.bigint "access_token_id"
  388. t.bigint "user_id", null: false
  389. t.bigint "web_push_subscription_id"
  390. t.index ["session_id"], name: "index_session_activations_on_session_id", unique: true
  391. t.index ["user_id"], name: "index_session_activations_on_user_id"
  392. end
  393. create_table "settings", force: :cascade do |t|
  394. t.string "var", null: false
  395. t.text "value"
  396. t.string "thing_type"
  397. t.datetime "created_at"
  398. t.datetime "updated_at"
  399. t.bigint "thing_id"
  400. t.index ["thing_type", "thing_id", "var"], name: "index_settings_on_thing_type_and_thing_id_and_var", unique: true
  401. end
  402. create_table "site_uploads", force: :cascade do |t|
  403. t.string "var", default: "", null: false
  404. t.string "file_file_name"
  405. t.string "file_content_type"
  406. t.integer "file_file_size"
  407. t.datetime "file_updated_at"
  408. t.json "meta"
  409. t.datetime "created_at", null: false
  410. t.datetime "updated_at", null: false
  411. t.index ["var"], name: "index_site_uploads_on_var", unique: true
  412. end
  413. create_table "status_pins", force: :cascade do |t|
  414. t.bigint "account_id", null: false
  415. t.bigint "status_id", null: false
  416. t.datetime "created_at", default: -> { "now()" }, null: false
  417. t.datetime "updated_at", default: -> { "now()" }, null: false
  418. t.index ["account_id", "status_id"], name: "index_status_pins_on_account_id_and_status_id", unique: true
  419. end
  420. create_table "statuses", id: :bigint, default: -> { "timestamp_id('statuses'::text)" }, force: :cascade do |t|
  421. t.string "uri"
  422. t.text "text", default: "", null: false
  423. t.datetime "created_at", null: false
  424. t.datetime "updated_at", null: false
  425. t.bigint "in_reply_to_id"
  426. t.bigint "reblog_of_id"
  427. t.string "url"
  428. t.boolean "sensitive", default: false, null: false
  429. t.integer "visibility", default: 0, null: false
  430. t.text "spoiler_text", default: "", null: false
  431. t.boolean "reply", default: false, null: false
  432. t.integer "favourites_count", default: 0, null: false
  433. t.integer "reblogs_count", default: 0, null: false
  434. t.string "language"
  435. t.bigint "conversation_id"
  436. t.boolean "local"
  437. t.bigint "account_id", null: false
  438. t.bigint "application_id"
  439. t.bigint "in_reply_to_account_id"
  440. t.boolean "local_only"
  441. t.index ["account_id", "id", "visibility", "updated_at"], name: "index_statuses_20180106", order: { id: :desc }
  442. t.index ["in_reply_to_id"], name: "index_statuses_on_in_reply_to_id"
  443. t.index ["reblog_of_id", "account_id"], name: "index_statuses_on_reblog_of_id_and_account_id"
  444. t.index ["uri"], name: "index_statuses_on_uri", unique: true
  445. end
  446. create_table "statuses_tags", id: false, force: :cascade do |t|
  447. t.bigint "status_id", null: false
  448. t.bigint "tag_id", null: false
  449. t.index ["status_id"], name: "index_statuses_tags_on_status_id"
  450. t.index ["tag_id", "status_id"], name: "index_statuses_tags_on_tag_id_and_status_id", unique: true
  451. end
  452. create_table "stream_entries", force: :cascade do |t|
  453. t.bigint "activity_id"
  454. t.string "activity_type"
  455. t.datetime "created_at", null: false
  456. t.datetime "updated_at", null: false
  457. t.boolean "hidden", default: false, null: false
  458. t.bigint "account_id"
  459. t.index ["account_id", "activity_type", "id"], name: "index_stream_entries_on_account_id_and_activity_type_and_id"
  460. t.index ["activity_id", "activity_type"], name: "index_stream_entries_on_activity_id_and_activity_type"
  461. end
  462. create_table "subscriptions", force: :cascade do |t|
  463. t.string "callback_url", default: "", null: false
  464. t.string "secret"
  465. t.datetime "expires_at"
  466. t.boolean "confirmed", default: false, null: false
  467. t.datetime "created_at", null: false
  468. t.datetime "updated_at", null: false
  469. t.datetime "last_successful_delivery_at"
  470. t.string "domain"
  471. t.bigint "account_id", null: false
  472. t.index ["account_id", "callback_url"], name: "index_subscriptions_on_account_id_and_callback_url", unique: true
  473. end
  474. create_table "tags", force: :cascade do |t|
  475. t.string "name", default: "", null: false
  476. t.datetime "created_at", null: false
  477. t.datetime "updated_at", null: false
  478. t.index "lower((name)::text) text_pattern_ops", name: "hashtag_search_index"
  479. t.index ["name"], name: "index_tags_on_name", unique: true
  480. end
  481. create_table "users", force: :cascade do |t|
  482. t.string "email", default: "", null: false
  483. t.datetime "created_at", null: false
  484. t.datetime "updated_at", null: false
  485. t.string "encrypted_password", default: "", null: false
  486. t.string "reset_password_token"
  487. t.datetime "reset_password_sent_at"
  488. t.datetime "remember_created_at"
  489. t.integer "sign_in_count", default: 0, null: false
  490. t.datetime "current_sign_in_at"
  491. t.datetime "last_sign_in_at"
  492. t.inet "current_sign_in_ip"
  493. t.inet "last_sign_in_ip"
  494. t.boolean "admin", default: false, null: false
  495. t.string "confirmation_token"
  496. t.datetime "confirmed_at"
  497. t.datetime "confirmation_sent_at"
  498. t.string "unconfirmed_email"
  499. t.string "locale"
  500. t.string "encrypted_otp_secret"
  501. t.string "encrypted_otp_secret_iv"
  502. t.string "encrypted_otp_secret_salt"
  503. t.integer "consumed_timestep"
  504. t.boolean "otp_required_for_login", default: false, null: false
  505. t.datetime "last_emailed_at"
  506. t.string "otp_backup_codes", array: true
  507. t.string "filtered_languages", default: [], null: false, array: true
  508. t.bigint "account_id", null: false
  509. t.boolean "disabled", default: false, null: false
  510. t.boolean "moderator", default: false, null: false
  511. t.bigint "invite_id"
  512. t.string "remember_token"
  513. t.string "chosen_languages", array: true
  514. t.index ["account_id"], name: "index_users_on_account_id"
  515. t.index ["confirmation_token"], name: "index_users_on_confirmation_token", unique: true
  516. t.index ["email"], name: "index_users_on_email", unique: true
  517. t.index ["reset_password_token"], name: "index_users_on_reset_password_token", unique: true
  518. end
  519. create_table "web_push_subscriptions", force: :cascade do |t|
  520. t.string "endpoint", null: false
  521. t.string "key_p256dh", null: false
  522. t.string "key_auth", null: false
  523. t.json "data"
  524. t.datetime "created_at", null: false
  525. t.datetime "updated_at", null: false
  526. t.bigint "access_token_id"
  527. t.bigint "user_id"
  528. t.index ["access_token_id"], name: "index_web_push_subscriptions_on_access_token_id"
  529. t.index ["user_id"], name: "index_web_push_subscriptions_on_user_id"
  530. end
  531. create_table "web_settings", force: :cascade do |t|
  532. t.json "data"
  533. t.datetime "created_at", null: false
  534. t.datetime "updated_at", null: false
  535. t.bigint "user_id", null: false
  536. t.index ["user_id"], name: "index_web_settings_on_user_id", unique: true
  537. end
  538. add_foreign_key "account_domain_blocks", "accounts", name: "fk_206c6029bd", on_delete: :cascade
  539. add_foreign_key "account_moderation_notes", "accounts"
  540. add_foreign_key "account_moderation_notes", "accounts", column: "target_account_id"
  541. add_foreign_key "accounts", "accounts", column: "moved_to_account_id", on_delete: :nullify
  542. add_foreign_key "admin_action_logs", "accounts", on_delete: :cascade
  543. add_foreign_key "backups", "users", on_delete: :nullify
  544. add_foreign_key "blocks", "accounts", column: "target_account_id", name: "fk_9571bfabc1", on_delete: :cascade
  545. add_foreign_key "blocks", "accounts", name: "fk_4269e03e65", on_delete: :cascade
  546. add_foreign_key "bookmarks", "accounts", on_delete: :cascade
  547. add_foreign_key "bookmarks", "statuses", on_delete: :cascade
  548. add_foreign_key "conversation_mutes", "accounts", name: "fk_225b4212bb", on_delete: :cascade
  549. add_foreign_key "conversation_mutes", "conversations", on_delete: :cascade
  550. add_foreign_key "custom_filters", "accounts", on_delete: :cascade
  551. add_foreign_key "favourites", "accounts", name: "fk_5eb6c2b873", on_delete: :cascade
  552. add_foreign_key "favourites", "statuses", name: "fk_b0e856845e", on_delete: :cascade
  553. add_foreign_key "follow_requests", "accounts", column: "target_account_id", name: "fk_9291ec025d", on_delete: :cascade
  554. add_foreign_key "follow_requests", "accounts", name: "fk_76d644b0e7", on_delete: :cascade
  555. add_foreign_key "follows", "accounts", column: "target_account_id", name: "fk_745ca29eac", on_delete: :cascade
  556. add_foreign_key "follows", "accounts", name: "fk_32ed1b5560", on_delete: :cascade
  557. add_foreign_key "identities", "users", on_delete: :cascade
  558. add_foreign_key "imports", "accounts", name: "fk_6db1b6e408", on_delete: :cascade
  559. add_foreign_key "invites", "users", on_delete: :cascade
  560. add_foreign_key "list_accounts", "accounts", on_delete: :cascade
  561. add_foreign_key "list_accounts", "follows", on_delete: :cascade
  562. add_foreign_key "list_accounts", "lists", on_delete: :cascade
  563. add_foreign_key "lists", "accounts", on_delete: :cascade
  564. add_foreign_key "media_attachments", "accounts", name: "fk_96dd81e81b", on_delete: :nullify
  565. add_foreign_key "media_attachments", "statuses", on_delete: :nullify
  566. add_foreign_key "mentions", "accounts", name: "fk_970d43f9d1", on_delete: :cascade
  567. add_foreign_key "mentions", "statuses", on_delete: :cascade
  568. add_foreign_key "mutes", "accounts", column: "target_account_id", name: "fk_eecff219ea", on_delete: :cascade
  569. add_foreign_key "mutes", "accounts", name: "fk_b8d8daf315", on_delete: :cascade
  570. add_foreign_key "notifications", "accounts", column: "from_account_id", name: "fk_fbd6b0bf9e", on_delete: :cascade
  571. add_foreign_key "notifications", "accounts", name: "fk_c141c8ee55", on_delete: :cascade
  572. add_foreign_key "oauth_access_grants", "oauth_applications", column: "application_id", name: "fk_34d54b0a33", on_delete: :cascade
  573. add_foreign_key "oauth_access_grants", "users", column: "resource_owner_id", name: "fk_63b044929b", on_delete: :cascade
  574. add_foreign_key "oauth_access_tokens", "oauth_applications", column: "application_id", name: "fk_f5fc4c1ee3", on_delete: :cascade
  575. add_foreign_key "oauth_access_tokens", "users", column: "resource_owner_id", name: "fk_e84df68546", on_delete: :cascade
  576. add_foreign_key "oauth_applications", "users", column: "owner_id", name: "fk_b0988c7c0a", on_delete: :cascade
  577. add_foreign_key "report_notes", "accounts", on_delete: :cascade
  578. add_foreign_key "report_notes", "reports", on_delete: :cascade
  579. add_foreign_key "reports", "accounts", column: "action_taken_by_account_id", name: "fk_bca45b75fd", on_delete: :nullify
  580. add_foreign_key "reports", "accounts", column: "assigned_account_id", on_delete: :nullify
  581. add_foreign_key "reports", "accounts", column: "target_account_id", name: "fk_eb37af34f0", on_delete: :cascade
  582. add_foreign_key "reports", "accounts", name: "fk_4b81f7522c", on_delete: :cascade
  583. add_foreign_key "session_activations", "oauth_access_tokens", column: "access_token_id", name: "fk_957e5bda89", on_delete: :cascade
  584. add_foreign_key "session_activations", "users", name: "fk_e5fda67334", on_delete: :cascade
  585. add_foreign_key "status_pins", "accounts", name: "fk_d4cb435b62", on_delete: :cascade
  586. add_foreign_key "status_pins", "statuses", on_delete: :cascade
  587. add_foreign_key "statuses", "accounts", column: "in_reply_to_account_id", name: "fk_c7fa917661", on_delete: :nullify
  588. add_foreign_key "statuses", "accounts", name: "fk_9bda1543f7", on_delete: :cascade
  589. add_foreign_key "statuses", "statuses", column: "in_reply_to_id", on_delete: :nullify
  590. add_foreign_key "statuses", "statuses", column: "reblog_of_id", on_delete: :cascade
  591. add_foreign_key "statuses_tags", "statuses", on_delete: :cascade
  592. add_foreign_key "statuses_tags", "tags", name: "fk_3081861e21", on_delete: :cascade
  593. add_foreign_key "stream_entries", "accounts", name: "fk_5659b17554", on_delete: :cascade
  594. add_foreign_key "subscriptions", "accounts", name: "fk_9847d1cbb5", on_delete: :cascade
  595. add_foreign_key "users", "accounts", name: "fk_50500f500d", on_delete: :cascade
  596. add_foreign_key "users", "invites", on_delete: :nullify
  597. add_foreign_key "web_push_subscriptions", "oauth_access_tokens", column: "access_token_id", on_delete: :cascade
  598. add_foreign_key "web_push_subscriptions", "users", on_delete: :cascade
  599. add_foreign_key "web_settings", "users", name: "fk_11910667b2", on_delete: :cascade
  600. end