Up until now, we have used Devise's Rememberable mechanism to re-log users
after the end of their browser sessions. This mechanism relies on a signed
cookie containing a token. That token was stored on the user's record,
meaning it was shared across all logged in browsers, meaning truly revoking
a browser's ability to auto-log-in involves revoking the token itself, and
revoking access from *all* logged-in browsers.
We had a session mechanism that dynamically checks whether a user's session
has been disabled, and would log out the user if so. However, this would only
clear a session being actively used, and a new one could be respawned with
the `remember_user_token` cookie.
In practice, this caused two issues:
- sessions could be revived after being closed from /auth/edit (security issue)
- auto-log-in would be disabled for *all* browsers after logging out from one
of them
This PR removes the `remember_token` mechanism and treats the `_session_id`
cookie/token as a browser-specific `remember_token`, fixing both issues.
* Add tests
* Add security-related tests
My first (unpublished) attempt at fixing the issues introduced (extremely
hard-to-exploit) security vulnerabilities, addressing them in a test.
* Fix authentication failures after going halfway through a sign-in attempt
* Refactor `authenticate_with_sign_in_token` and `authenticate_with_two_factor` to make the two authentication steps more obvious
* 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 tests
* Fix serialization of followers/following counts when user hides their network
Fixes#16382
Signed-off-by: Claire <claire.github-309c@sitedethib.com>
* Fix followers synchronization mechanism not working when URI has empty path
To my knowledge, there is no current implementation on the fediverse
that can use bare domains (e.g., actor is at https://example.org instead of
something like https://example.org/actor) that also plans to support the
followers synchronization mechanism. However, Mastodon's current implementation
would exclude such accounts from followers list.
Also adds tests and rename them to reflect the proper method names.
* Move url prefix regexp to its own constant
* 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
An uncommon but somewhat difficult to digagnose issue is dealing with
improperly-seeded databases. In such cases, instance-signed fetches will
fail with a ActiveRecord::RecordNotFound error, usually caught and handled
as generic 404, leading people to think the remote resource itself has not
been found, while it's the local instance actor that does not exist.
This commit changes the code so that failure to find the instance actor
automatically creates a new one, so that improperly-seeded databases do
not cause any issue.
* Fix media processing getting stuck on too much stdin/stderr
See thoughtbot/terrapin#5
* Remove dependency on paperclip-av-transcoder gem
* Remove dependency on streamio-ffmpeg gem
* Disable stdin on ffmpeg process
Database serialization failure occurs when a read-replica is used
and a query takes long enough that rows on the primary database
become unavailable. It should return HTTP 503 as it is temporary.
Re-order rescue definitions according to their status codes
* Fix crash on receiving requests with missing Digest header
Return an error pointing out that Digest is missing, instead of crashing.
Fixes#15743
* Fix from review feedback
* Fix URI of repeat follow requests not being recorded
In case we receive a “repeat” or “duplicate” follow request, we automatically
fast-forward the accept with the latest received Activity `id`, but we don't
record it.
In general, a “repeat” or “duplicate” follow request may happen if for some
reason (e.g. inconsistent handling of Block or Undo Accept activities, an
instance being brought back up from the dead, etc.) the local instance thought
the remote actor were following them while the remote actor thought otherwise.
In those cases, the remote instance does not know about the older Follow
activity `id`, so keeping that record serves no purpose, but knowing the most
recent one is useful if the remote implementation at some point refers to it
by `id` without inlining it.
* Add tests
This also skips fetching the actor completely.
This will be useful if we end up distributing Update activities linked to
account suspensions more widely (they are currently only delivered to
the suspended account's followers), as currently, instances not knowing
about the suspended account would fetch it to then process the suspension.
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
Unlike locally-issued blocks, they weren't clearing follow
relationships in both directions, follow requests or notifications.
Co-authored-by: Claire <claire.github-309c@sitedethib.com>
* 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
* 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>
