This was introduced to let the functionality happen
via mmctl, but it does not differentiate between
sysadmins vs people having physical access to machines.
Therefore, we remove it and only keep the local endpoint
accessible from mmctl --local.
https://mattermost.atlassian.net/browse/MM-41184
```release-note
The api/v4/config/migrate API endpoint has been removed
in favor of the mmctl --local endpoint. API clients
won't be able to access this endpoint without having
physical access to the server.
```
Before gorp, to use sqlx we had to resort to using
internal structs without the `db:"-"` tag so that
it doesn't interfere with the gorp table generation.
Now that it's gone, we can use the original model structs.
```release-note
NONE
```
We were noting down the time of execution of a function and
expected that to match with the exact interval that we set.
However, sometimes in very busy CI environments, the goroutine
does not get a chance to be scheduled at the right time for
the interval to match exactly. Sometimes, we see a delay of
1 second.
To account for this, we increase the interval to 3 seconds,
and consider a 1 second buffer for this. If we see lags of
more than a second, then this test becomes useless and we
would need to remove it entirely.
https://mattermost.atlassian.net/browse/MM-41209
```release-note
NONE
```
* MM-41066: fixes replies when binary_parameters=yes
Migrating post_store from gorp to sqlx breaks replies on threads.
Why? Participants is a jsonb field and when binary_parameters is set to
'yes' it is failing to insert because it needs the version (1) to be
prepended to the bytes array ([]byte{0x01}).
The easiest fix on this is to cast to a string when inserting.
There is a drawback though, non utf8 characters would be replaced by the
question mark icon �.
This commit does exactly that, casts StringArray to a string.
* Adds a comment
* Removes unnecessary conversion
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* add Boards to DataRetention, add hook for data retention
* remove replaces
* update hook to remove parameter
* add boards data retention to telemetry
* fix unit test
* update test, update hooks
* update RunDataRetention server version
* put behind a feature flag
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* Update morph dependency to use a newer version
* remove timeout check for migrations statements
* store/sqlstore: reset timeout for mysql while creating db for migrations
* update morph to v0.2.1
- Fixed the Makefile command so that it can be run in CI
- Changed to go install to make it compatible with >1.16 versions.
- Removed unnecessary lines in go.tools.mod now that we aren't using that.
```release-note
NONE
```
* Add playbooks related permissions.
* Add RolesGrantPermission to pluginapi
* Fixing scopes.
* New defaults.
* Fix defaults
* Fix tests.
* Fix migration.
* More test and migration fixes.
* Need to add everything to system admin too.
* Move to 63
* Feedback fixes.
* Fix system manager editing playbook permissions.
* Enable receiving binary websocket messages
* Improve error message
* Prefer anonymous declaration
* Simplify
* Improve test
* Use MessagePack to clone WebSocketRequest struct
* Use short form
* Fix test
* MM-36862: removes participant from thread
Removing a participant upon last reply deleted from thread didn't work
reliably, a suspect on this is the replica lag, since we are first
deleting the post and then counting non-deleted posts of the participant
to decide on whether to delete or not.
The findings that led to this conclusion is that the reply count gets
updated but the participant is not removed (participant removal depends
on the number of replies this participant has in the thread.)
This commit fixes that by removing first the participant and then
deleting the post.
So we delete the participant if they have 1 post in that thread, and
then we delete the post, so now they have no posts in the thread.
* Makes deleting posts transactional
This commit makes deleting a post transactional and also tries to fix
permanent deletion of posts.
Currently when we permanently delete all posts by a user we don't update
the threads reply count nor the participant's array. This commit tries
to fix that.
* Adds comments on deleting posts
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* MM-38674 - AB test for first guided channel creation
* set default value for the guided creation as false
Co-authored-by: Pablo Velez Vidal <pablo.velez@mattermost.com>
Summary
The addition of the TotalMsgCountRoot and MsgCountRoot columns to support CRT caused several issues with previously read threads and channels being marked as unread. Previously we attempted to fix this purely in a SQL migration [MM-35345][MM-35494] fixes for incorrect mentions and unreads for threads and channels #17803 but that turned out to be too heavy and it was decided to break up some of the fixes into async jobs.
This PR implements an async job to mark channels as read if there are no user posts since the last time the user viewed the channel.
Ticket Link
https://mattermost.atlassian.net/browse/MM-37013
* MM-39612: Make acquiring and removing connections atomic
The reconnect phase of a websocket was split into two parts:
one where we check if a connection with a given connectionID
exists or not. And second, where we remove that connection
and insert the new connection again in the index.
This would lead to a race where it would be possible
for 2 concurrent requests for the same connectionID to go through
which would lead to separate goroutines working on the same dead queue.
We simplify this by removing the connection from the index
in the check connection stage itself. And then just add that
during register phase.
And to distinguish between a fresh and an old connection, we add
a new field called reuseCount.
While here, we also cleanup some old comments and add more
in some places.
https://mattermost.atlassian.net/browse/MM-39612
```release-note
NONE
```
* remove unused method
```release-note
NONE
```
* race test
```release-note
NONE
```
