There is an idempotency key generated by clients when authoring a post,
and stored in Redis, to ensure that if a user or client retries posting
the same status, we don't get a duplicate.
Hachyderm.io has been experiencing some filesystem and database
performance issues, causing database writes to be slow. This can mean
that there are successful posts, but the reverse proxy returns 504
Gateway Timeout before the idempotency status has been updated; users or
clients who retry (such as Tusky which retries automatically, see
tuskyapp/Tusky#2951) can re-try the same post with the same idempotency
key before it has actually been recorded in Redis, leading to duplicate
posts.
To address this issue, move all of the database updates after the
initial transaction that creates the status into the
`postprocess_status!` method, so we can insert the idempotency key
immediately after the status has been created, significantly reducing
the window in which the status could be created but the idempotency key
not yet stored.
Note: this has not yet been tested; I'm submitting this PR for
discussion and to offer to the Hachyderm.io admins to try out to fix the
multiple posting problem.
Co-authored-by: Brian Campbell <brcampbell@beta.team>
* Fix trying to fetch posts from other users when fetching featured posts
* Rate-limit discovery of new subdomains
* Put a limit on recursively discovering new accounts
Before this change, the following error would cause VerifyAccountLinksWorker to fail:
NoMethodError: undefined method `downcase' for nil:NilClass
[PROJECT_ROOT]/app/services/verify_link_service.rb:31 :in `block in link_back_present?`
- Change verification to happen in `default` queue
- Change verification worker to only be queued if there's something to do
- Add `link` tags from metadata fields to page header of profiles
* Fix followers count not being updated when migrating follows
Fixes#19900
* Fix validation error in SynchronizeFeaturedTagsCollectionWorker
Also saves remote user's chosen case for hashtags
* Limit remote featured tags before validation
- Only exact search matches for queries with < 5 characters
- Do not support queries with `offset` (pagination)
- Return HTTP 401 on truthy `resolve` instead of overriding to false
* Change featured tag updates to add/remove activity
* Fix to check for the existence of feature tag
* Rename service and worker
* Merge AddHashtagSerializer with AddSerializer
* Undo removal of sidekiq_options
* Add LIMIT of featured tag to instance API response
* Add featured_tags_collection_url to Account
* Add synchronization of remote featured tags
* Deliver update activity when updating featured tag
* Remove featured_tags_collection_url
* Revert "Add featured_tags_collection_url to Account"
This reverts commit cff349fc27.
* Add hashtag sync from featured collections
* Fix tag name normalize
* Add target option to fetch featured collection
* Refactor fetch_featured_tags_collection_service
* Add LIMIT of featured tag to v1/instance API response
* Move ActivityPub::FetchRemoteAccountService to ActivityPub::FetchRemoteActorService
ActivityPub::FetchRemoteAccountService is kept as a wrapper for when the actor is
specifically required to be an Account
* Refactor SignatureVerification to allow non-Account actors
* fixup! Move ActivityPub::FetchRemoteAccountService to ActivityPub::FetchRemoteActorService
* Refactor ActivityPub::FetchRemoteKeyService to potentially return non-Account actors
* Refactor inbound ActivityPub payload processing to accept non-Account actors
* Refactor inbound ActivityPub processing to accept activities relayed through non-Account
* Refactor how Account key URIs are built
* Refactor Request and drop unused key_id_format parameter
* Rename ActivityPub::Dereferencer `signature_account` to `signature_actor`
* Add a more descriptive PrivateNetworkAddressError exception class
* Remove unnecessary exception class to rescue clause
* Remove unnecessary include to JsonLdHelper
* Give more neutral error message when too many webfinger redirects
* Remove unnecessary guard condition
* Rework how “ActivityPub::FetchRemoteAccountService” handles errors
Add “suppress_errors” keyword argument to avoid raising errors in
ActivityPub::FetchRemoteAccountService#call (default/previous behavior).
* Rework how “ActivityPub::FetchRemoteKeyService” handles errors
Add “suppress_errors” keyword argument to avoid raising errors in
ActivityPub::FetchRemoteKeyService#call (default/previous behavior).
* Fix Webfinger::RedirectError not being a subclass of Webfinger::Error
* Add suppress_errors option to ResolveAccountService
Defaults to true (to preserve previous behavior). If set to false,
errors will be raised instead of caught, allowing the caller to be
informed of what went wrong.
* Return more precise error when failing to fetch account signing AP payloads
* Add tests
* Fixes
* Refactor error handling a bit
* Fix various issues
* Add specific error when provided Digest is not 256 bits of base64-encoded data
* Please CodeClimate
* Improve webfinger error reporting
* Change unapproved and unconfirmed account to not be accessible in the REST API
* Change Account#searchable? to reject unconfirmed and unapproved users
* Disable search for unapproved and unconfirmed users in Account.search_for
* Disable search for unapproved and unconfirmed users in Account.advanced_search_for
* Remove unconfirmed and unapproved accounts from Account.searchable scope
* Prevent mentions to unapproved/unconfirmed accounts
* Fix some old tests for Account.advanced_search_for
* Add some Account.advanced_search_for tests for existing behaviors
* Add some tests for Account.search_for
* Add Account.advanced_search_for tests unconfirmed and unapproved accounts
* Add Account.searchable tests
* Fix Account.without_unapproved scope potentially messing with previously-applied scopes
* Allow lookup of unconfirmed/unapproved accounts through /api/v1/accounts/lookup
This is so that the API can still be used to check whether an username is free
to use.
* Change e-mail notifications to only be sent when recipient is offline
Change the default for follow and mention notifications back on
* Add preference to always send e-mail notifications
* Change wording
* Change post text edit to not be considered significant if it's identical after reformatting
* We don't need to clear previous change information anymore
* Require status edits to be explicit, except for poll tallies
* Fix tests
* Add some tests
* Add poll-related tests
* Add HTML-formatting related tests
Not completely sure this could actually have any ill effect, but if
`RemoveStatusService` fails to acquire a lock in an
`ActivityPub::ProcessingWorker` job processing a `Delete`, the status
is currently discarded and causes a job failure but the next time the
job is attempted, it will skip deleting the status due to it being
discarded.
This commit makes the behavior of `RemoveStatusService` a bit more consistent
in case of failure to acquire the lock.
* Optimize and clean up `response_to_recipient?` CTE
Marginally improve performances, and make the CTE much more readable
* Limit max depth in `response_to_recipient?` CTE
* Fix edits with no actual changes being allowed locally
* Fix edits with no actual changes being allowed through ActivityPub
* Fix false positive changes caused by description processing in model
* Fix not recording poll expiration update
* Fix test
* Revert changes to ProcessStatusUpdateService
* Various fixes and improvements
* Fix code style issues
* Various changes and improvements
* Add guard clause
* Fix searching for an already-known status by URL not working
* Fix Update processing from statuses prior to 20220302232632
`ordered_media_attachment_ids_changed?` would return `true` when going from
`nil` to anything (including `[]`).
* Add tests
* Change how changes to media attachments are stored for edits
Fix not being able to re-order media attachments
* Fix not broadcasting updates when polls/media is changed through ActivityPub
* Various fixes and improvements
* Update app/models/report.rb
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Add tracking of media attachment description changes
* Change poll in status edit to have a structure closer to the real one
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Add trending statuses
* Fix dangling items with stale scores in localized sets
* Various fixes and improvements
- Change approve_all/reject_all to approve_accounts/reject_accounts
- Change Trends::Query methods to not mutate the original query
- Change Trends::Query#skip to offset
- Change follow recommendations to be refreshed in a transaction
* Add tests for trending statuses filtering behaviour
* Fix not applying filtering scope in controller
* Add appeals
* Add ability to reject appeals and ability to browse pending appeals in admin UI
* Add strikes to account page in settings
* Various fixes and improvements
- Add separate notification setting for appeals, separate from reports
- Fix style of links in report/strike header
- Change approving an appeal to not restore statuses (due to federation complexities)
- Change style of successfully appealed strikes on account settings page
- Change account settings page to only show unappealed or recently appealed strikes
* Change appealed_at to overruled_at
* Fix missing method error
* Add notifications when a reblogged status has been updated
* Change wording to say "edit" instead of "update" and add missing controls
* Replace previous update notifications with the most up-to-date one
* Add editing for published statuses
* Fix change of multiple-choice boolean in poll not resetting votes
* Remove the ability to update existing media attachments for now
* Fix Sidekiq warnings about JSON serialization
This occurs on every symbol argument we pass, and every symbol key in hashes,
because Sidekiq expects strings instead.
See https://github.com/mperham/sidekiq/pull/5071
We do not need to change how workers parse their arguments because this has
not changed and we were already converting to symbols adequately or using
`with_indifferent_access`.
* Set Sidekiq to raise on unsafe arguments in test mode
In order to more easily catch issues that would produce warnings in production
code.
* Add support for editing for published statuses
* Fix references to stripped-out code
* Various fixes and improvements
* Further fixes and improvements
* Fix updates being potentially sent to unauthorized recipients
* Various fixes and improvements
* Fix wrong words in test
* Fix notifying accounts that were tagged but were not in the audience
* Fix mistake
* Add trending links
* Add overriding specific links trendability
* Add link type to preview cards and only trend articles
Change trends review notifications from being sent every 5 minutes to being sent every 2 hours
Change threshold from 5 unique accounts to 15 unique accounts
* Fix tests
* Add tests
* Fix some link previews being incorrectly generated from different prior links
PR #12403 added a cache to avoid redundant queries when the OEmbed endpoint can
be guessed from the URL. This caching mechanism is not perfectly correct as
there is no guarantee that all pages from a given domain share the same
OEmbed provider endpoint.
This PR prevents the FetchOEmbedService from caching OEmbed endpoint that
cannot be generalized by replacing a fully-qualified URL from the endpoint's
parameters, greatly reducing the number of incorrect cached generalizations.
* Add support for fetching Create and Announce activities by URI
This should improve compatibility with ZAP and offer a way to fetch boosts,
which is currently not possible.
* Add tests
* Add followed_by? to account_interactions
* Add RemoveFromFollowersService
* Fix AccountBatch to use RemoveFromFollowersService
* Add remove from followers API
Eliminate need to have custom notifications filtering logic in the
streaming API code by publishing notifications into a separate stream
and then simply using the multi-stream capability to subscribe to that
stream when necessary
Fix issue #16603 undefined method `serialize_payload' for Unsuspend Account Service error.
It seems that this service forgot to `include Payloadable` so that `serialize_payload` could not be found in this service.
* Add account statuses cleanup policy model
* Record last inspected toot to delete to speed up successive calls to statuses_to_delete
* Add service to cleanup a given account's statuses within a budget
* Add worker to go through account policies and delete old toots
* Fix last inspected status id logic
All existing statuses older or equal to last inspected status id must be
kept by the current policy. This is an invariant that must be kept so that
resuming deletion from the last inspected status remains sound.
* Add tests
* Refactor scheduler and add tests
* Add user interface
* Add support for discriminating based on boosts/favs
* Add UI support for min_reblogs and min_favs, rework UI
* Address first round of review comments
* Replace Snowflake#id_at_start with with_random parameter
* Add tests
* Add tests for StatusesCleanupController
* Rework settings page
* Adjust load-avoiding mechanisms
* Please CodeClimate
* Add account_notes relationship
* Add tests
* Fix owned account notes not being deleted when an account is deleted
* Add post-migration to clean up orphaned account notes
* Fix Delete and Create-related locks expiring too fast
Fixes#16238
By default, RedisLock expires after 10 seconds, which may not be enough to
process statuses, especially when those have attached media files.
This commit extends those 10 seconds to 15 minutes, which should be plenty
enough to handle any status, while being short enough to not waste many
sidekiq job retries in the exceedingly rare case in which a sidekiq process
would crash when processing a `Create` or `Delete`.
* Fix other RedisLock autorelease durations
Fixes#15645
- things that only perform a few simple database queries (e.g. finding and
saving a record) have been left unchanged, so they'll still use the default
10s duration
- things that perform significantly more complex database queries have been
changed to a 5 minutes timeout
- things that perform multiple HTTP queries have been changed to a 15 minutes
timeout
If a status with a hashtag becomes very popular, it stands to
reason that the hashtag should have a chance at trending
Fix no stats being recorded for hashtags that are not allowed
to trend, and stop ignoring bots
Remove references to hashtags in profile directory from the code
and the admin UI
* Add tests
* Ensure deleted statuses are marked as such
* Save some redis memory by not storing URIs in delete_upon_arrival values
* Avoid possible race condition when processing incoming Deletes
* Avoid potential duplicate Delete forwards
* Lower lock durations to reduce issues in case of hard crash of the Rails process
* Check for `lock.aquired?` and improve comment
* Refactor RedisLock usage in app/lib/activitypub
* Fix using incorrect or non-existent sender for relaying Deletes
* Update devise-two-factor to unreleased fork for Rails 6 support
Update tests to match new `rotp` version.
* Update nsa gem to unreleased fork for Rails 6 support
* Update rails to 6.1.3 and rails-i18n to 6.0
* Update to unreleased fork of pluck_each for Ruby 6 support
* Run "rails app:update"
* Add missing ActiveStorage config file
* Use config.ssl_options instead of removed ApplicationController#force_ssl
Disabled force_ssl-related tests as they do not seem to be easily testable
anymore.
* Fix nonce directives by removing Rails 5 specific monkey-patching
* Fix fixture_file_upload deprecation warning
* Fix yield-based test failing with Rails 6
* Use Rails 6's index_with when possible
* Use ActiveRecord::Cache::Store#delete_multi from Rails 6
This will yield better performances when deleting an account
* Disable Rails 6.1's automatic preload link headers
Since Rails 6.1, ActionView adds preload links for javascript files
in the Links header per default.
In our case, that will bloat headers too much and potentially cause
issues with reverse proxies. Furhermore, we don't need those links,
as we already output them as HTML link tags.
* Switch to Rails 6.0 default config
* Switch to Rails 6.1 default config
* Do not include autoload paths in the load path
* Update twitter-text from 1.14 to 3.1.0
* Disable emoji parsing
* Properly depend on twitter-text for url detection
* Fix some URLs being wrongly detected client-side
* Add test for server-side validation of non-autolinkable URLs
* Fix server-side status length counting
* Change ResolveAccountService's handling of skip_webfinger
Change it so it never makes any webfinger query, as the name would imply.
* Add tests
* Change FollowService to not take an URI for target_account
* Restore domain-block check in FollowService
* Fix tests
* disable NewCops
* update TargetRubyVersion
* Fix Lint/MissingSuper for ActiveModelSerializers::Model
* Fix Lint/MissingSuper for feed
* Fix Lint/FloatComparison
* Do not use instance variables
* Fix being able to import more than allowed number of follows
Without this commit, if someone tries importing a second list of accounts to
follow before the first one has been processed, this will queue imports for
the two whole lists, even if they exceed the account's allowed number of
outgoing follows.
This commit changes it so the individual queued imports aren't exempt from
the follow limit check (they remain exempt from the rate-limiting check
though).
* Catch validation errors to not re-queue failed follows
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
As a regression from the recent optimizations, mentions were left untouched
until `account.destroy`, which would then delete them individually,
and executing queries to find and delete associated notifications, resulting
in a massive slowdown.
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Delete status records by batches of 50
* Do not precompute values that are only used once
* Do not generate redis events for removal of public toots older than two weeks
* Filter reported toots a priori for polls and status deletion
* Do not process reblogs when cleaning up public timelines
As in Mastodon proper, reblogs don't appear in public TLs
* Clean the deleted account's own feed in one go
* Refactor Account#clean_feed_manager and List#clean_feed_manager
* Delete instead of destroy a few more associations
* Fix preloading
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Fix deleting polls not deleting notifications
* Fix fav notification deletion when deleting a toot
* Refactor DeleteAccountService spec
* Add DeleteAccountService tests for other associations and notifications
* Add favourite handling spec in status removal
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Fix account deletion workers being queued multiple times for a single account
* Fix poll votes being unnecessarily instantiated on poll deletion
* Fix favourites being unnecessarily instantiated on status deletion
* Remove inaccurate comments
* Delete polls instead of destroying them
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Fix ResolveAccountService accepting mismatching acct: URI
* Set attributes that should be updated regardless of suspension
* Fix key fetching
* Automatically merge remote accounts with duplicate `uri`
* Add tests
* Add "tootctl accounts fix-duplicates"
Finds duplicate accounts sharing a same ActivityPub `id`, re-fetch them and
merge them under the canonical `acct:` URI.
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Improve searching for private toots from URL
Most of the time, when sharing toots, people use the toot URL rather than
the toot URI, which makes sense since it is the user-facing URL.
In Mastodon's case, the URL and URI are different, and Mastodon does not
have an index on URL, which means searching a private toot by URL is done
with a slow query that will only succeed for very recent toots.
This change gets rid of the slow query, and attempts to guess the URI from
URL instead, as Mastodon's are predictable.
* Add tests
* Only return status with guessed uri if url matches
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* Add indication to admin UI of whether a report has been forwarded
* Rework how forwarded status is displayed
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
Extract logic for determining ActivityPub inboxes to send deletes
to to its own class and explicitly include the person the status
replied to (even if not mentioned), people who favourited it, and
people who replied to it (though that one is still not recursive)
* Fix webfinger redirect handling in ResolveAccountService
ResolveAccountService#process_webfinger! handled a one-step webfinger
redirection, but only accepting the result if it matched the exact URI passed
as input, defeating the point of a redirection check.
Instead, use the same logic as in `ActivityPub::FetchRemoteAccountService`,
updating the resulting `acct:` URI with the result of the first webfinger
query.
* Add tests
Nginx can be configured to bypass proxy cache when a special header
is in the request. If the response is cacheable, it will replace
the cache for that request. Proxy caching of media files is
desirable when using object storage as a way of minimizing bandwidth
costs, but has the drawback of leaving deleted media files for
a configured amount of cache time. A cache buster can make those
media files immediately unavailable. This especially makes sense
when suspending and unsuspending an account.
* Fix crash in SuspendAccountWorker
`follows` is an array thanks to `to_a`
* Fix code style issue
Co-authored-by: Eugen Rochko <eugen@zeonfederated.com>
e.g. if someone on an instance that previously had followers gets mentioned
in a private toot, before this PR, they would not receive a
Collection-Synchronization header and may show the toot to the former followers
in addition to the mentioned person.
* Add support for followers synchronization on the receiving end
Check the `collectionSynchronization` attribute on `Create` and `Announce`
activities and synchronize followers from provided collection if possible.
* Add tests for followers synchronization on the receiving end
* Add support for follower synchronization on the sender's end
* Add tests for the sending end
* Switch from AS attributes to HTTP header
Replace the custom `collectionSynchronization` ActivityStreams attribute by
an HTTP header (`X-AS-Collection-Synchronization`) with the same syntax as
the `Signature` header and the following fields:
- `collectionId` to specify which collection to synchronize
- `digest` for the SHA256 hex-digest of the list of followers known on the
receiving instance (where “receiving instance” is determined by accounts
sharing the same host name for their ActivityPub actor `id`)
- `url` of a collection that should be fetched by the instance actor
Internally, move away from the webfinger-based `domain` attribute and use
account `uri` prefix to group accounts.
* Add environment variable to disable followers synchronization
Since the whole mechanism relies on some new preconditions that, in some
extremely rare cases, might not be met, add an environment variable
(DISABLE_FOLLOWERS_SYNCHRONIZATION) to disable the mechanism altogether and
avoid followers being incorrectly removed.
The current conditions are:
1. all managed accounts' actor `id` and inbox URL have the same URI scheme and
netloc.
2. all accounts whose actor `id` or inbox URL share the same URI scheme and
netloc as a managed account must be managed by the same Mastodon instance
as well.
As far as Mastodon is concerned, breaking those preconditions require extensive
configuration changes in the reverse proxy and might also cause other issues.
Therefore, this environment variable provides a way out for people with highly
unusual configurations, and can be safely ignored for the overwhelming majority
of Mastodon administrators.
* Only set follower synchronization header on non-public statuses
This is to avoid unnecessary computations and allow Follow-related
activities to be handled by the usual codepath instead of going through
the synchronization mechanism (otherwise, any Follow/Undo/Accept activity
would trigger the synchronization mechanism even if processing the activity
itself would be enough to re-introduce synchronization)
* Change how ActivityPub::SynchronizeFollowersService handles follow requests
If the remote lists a local follower which we only know has sent a follow
request, consider the follow request as accepted instead of sending an Undo.
* Integrate review feeback
- rename X-AS-Collection-Synchronization to Collection-Synchronization
- various minor refactoring and code style changes
* Only select required fields when computing followers_hash
* Use actor URI rather than webfinger domain in synchronization endpoint
* Change hash computation to be a XOR of individual hashes
Makes it much easier to be memory-efficient, and avoid sorting discrepancy issues.
* Marginally improve followers_hash computation speed
* Further improve hash computation performances by using pluck_each
There are edge cases where requests to certain hosts timeout when
using the vanilla HTTP.rb gem, which the goldfinger gem uses. Now
that we no longer need to support OStatus servers, webfinger logic
is so simple that there is no point encapsulating it in a gem, so
we can just use our own Request class. With that, we benefit from
more robust timeout code and IPv4/IPv6 resolution.
Fix#14091
* Add bell button
Fix#4890
* Remove duplicate type from post-deployment migration
* Fix legacy class type mappings
* Improve query performance with better index
* Fix validation
* Remove redundant index from notifications
DISTINCT clause removes duplicated records according to all the selected
attributes. In reality, it can remove duplicated records only looking at
statuses.id, but the clause confuses the query planner and yields
insufficient performance.
The behavior is also problematic if the scope produced by HashQueryService
is used to query columns without id (using pluck method, for example). The
scope is expected to contain unique statuses, but the uniquness will be
evaluated with some arbitrary columns other than id.
GROUP BY clause resolves those problem by explicitly specifying the
column to take into account for the record distinction.
A workaround for the problem of DISTINCT clause in
Api::V1::Timelines::TagController is no longer necessary and removed.
- Change audio files to not be stripped of metadata
- Automatically extract cover art from audio if it exists
- Add `thumbnail` parameter to `POST /api/v1/media`, `POST /api/v2/media` and `PUT /api/v1/media/:id`
- Add `icon` to represent it in attachments in ActivityPub
- Fix `preview_url` containing URL of missing missing image when there is no thumbnail instead of null
- Fix duration of audio not being displayed on public pages until the file is loaded
* Split media cleanup from reject-media domain blocks to its own service
* Slightly improve ClearDomainMediaService error handling
* Lower DomainClearMediaWorker to lowest-priority queue
* Do not catch ActiveRecord::RecordNotFound in domain block workers
* Fix DomainBlockWorker spec labels
* Add some specs
* Change domain blocks to immediately mark accounts as suspended
Rather than doing so sequentially, account after account, while cleaning
their data. This doesn't change much about the time the block takes to
complete, but it immediately prevents interaction with the blocked domain,
while up to now, it would only be guaranteed when the process ends.
Mastodon enforces the “sensitive” flag on media attachments whenever a toot
is posted with a Content Warning. However, it does so *after* potentially
converting the Content Warning to toot text (when there is no toot text),
which leads to inconsistent and surprising behavior for API clients.
This commit fixes this inconsistency.
