* Migration completed
* Fix tests
* Reduce to one line
* Fix: change to plain error
* Fix imports
* Trigger CI
* Fix i18
* Fix merge with master
* Trigger CI
MM-27744 disable Zap for unit tests.
Zap has no concept of shutdown or close. Zap is only shutdown when the app exits. Not a problem for console logging, but when creating a new Zap logger that outputs to files on every unit test, that leaves no easy way to clean up until process exit. Depending on what else is running this can exhaust all file handles and cause unit tests to fail.
Zap is now disabled unit tests and uses Logr instead, regardless of config settings. `make test-server` peak file handle usage dropped from ~5K to less than 100.
GetPostsSince is used when loading posts for a channel.
An opportunity for optimization is that the primary SQL query
is repeated twice and then a UNION is constructed for the results.
```
SELECT
*
FROM
Posts
WHERE
UpdateAt > :Time AND ChannelId = :ChannelId
LIMIT 1000
```
But we can use a CTE for this which caches the results to be reused later.
This leads to the main query being executed once rather than twice. And from
Postgres 12 onwards, CTEs can be inlined which opens the door to further optimizations.
From the docs (https://www.postgresql.org/docs/10/queries-with.html)
> A useful property of WITH queries is that they are evaluated only once per
execution of the parent query, even if they are referred to more than once
by the parent query or sibling WITH queries. Thus, expensive calculations
that are needed in multiple places can be placed within a WITH query to avoid redundant work.
Another possible application is to prevent unwanted multiple evaluations of functions
with side-effects. However, the other side of this coin is that the optimizer is
less able to push restrictions from the parent query down into a WITH query than an ordinary subquery.
In our case, the caveat does not apply because we are only filtering columns and not rows,
so we can safely use it.
Following are the query plan comparisons:
Old: http://tatiyants.com/pev/#/plans/plan_1599394993105
New: http://tatiyants.com/pev/#/plans/plan_1599395970886
As we can see, in old bitmap index scan+heap scan happens twice, but in the new one,
it happens only once.
This has been load tested with a large dataset and confirmed to exhibit good improvements.
We got hit by yet another random text search issue, but this was
on the more extreme end because this was a string of length 4 and still
it collided.
I am not sure if there is a fool proof solution to this than just
going for a larger string.
https://mattermost.atlassian.net/browse/MM-27953
* MM-23063: Fix flaky test TestTermsOfServiceStore
The call to testGetLatestTermsOfService would happen after testSaveTermsOfService
which would persist the data between calls. Therefore, if they were to happen under
a milisecond, there would be 2 rows with the same CreateAt timestamp and the DB
would randomly return any row.
If this were to happen, then the wrong row would not match the user id and would fail.
To fix this, we just clear the table data before proceeding with the test.
https://mattermost.atlassian.net/browse/MM-23063
* Use t.Cleanup to refactor things
* Fix cleanup of tables
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* Fix return nil
A nil error is an interface having nil value but a non-nil type.
This leads to unexpected errors in the code well-documented here:
https://golang.org/doc/faq#nil_error
We fix this by passing an additional parameter to handle it.
Caught using https://github.com/dgryski/semgrep-go
* Bring back auto-generated line
* MM-23832: Initial set of changes
* MM-23832: further iteration
* MM-23832: further iteration
* MM-23832: further iteration
* MM-23832: Fixes merge.
* create migration for new Roles
* MM-23832: Renames some roles.
* MM-23832: Adds ability to see logs.
* MM-23832: Removes manage roles from restricted admin.
* MM-23832: Make authentication section read-only for restricted admin.
* MM-23832: Allow restricted admin to purge caches.
* MM-23832: Adds ability to recycle DB connections.
* MM-23832: Adds ability to purge indexes.
* MM-23832: Adds ability to test email and S3 config.
* MM-23832: Adds abilituy to read job status.
* MM-23832: Adds ability to read plugin statuses.
* MM-23832: Renames Restricted Admin to System Manager.
* MM-23832: Adds manage team roles to system_user_manager.
* MM-23832: Updates some permissions.
* MM-23832: Allow get all channels and get moderations.
* MM-23832: Adds some permissions to User Manager.
* MM-23832: Remove write users from user manager.
* MM-23832: Changes permissions for the usermanagement > users sysconsole section.
* MM-23832: Removes read_settings and write_settings permissions. Ensures the usermanagement parent permissions encompass the sub-permissions.
* MM-23832: Updates permissions.
* MM-23832: Changes some permissions checks, adds new permissions to roles.
* MM-23832: Adds ability to update a role.
* MM-23832: Permissions updates.
* MM-23832: Removes write access to plugins for system manager.
* MM-23832: Removes read compliance from new roles.
* MM-23832: Adds mock for new roles creation migration.
* MM-23832: Changes to variadic param.
* MM-23832: Removes some duplication in the permissions model. Renames some permissions constants.
* MM-23832: Updates some migrations.
* MM-23832: Removes some unnecessary constants.
* MM-23832: Changes back to old app method name.
* MM-23832: Fixes incorrect permission check.
* MM-23832: Changes write to read permission check.
* MM-23832: Removes the authentication permission from link/unlink group.
* MM-23832: Enable testing LDAP with read permissions.
* MM-23832: Make testing elasticsearch a read permission.
* MM-23832: Warn metrics are associated to any system console read permissions.
* MM-23832: Updates some permissions checks.
* MM-23832: Removes non-systemconsole permissions from roles.
* MM-23832: Update default permission assignment of sysadmin.
* MM-23832: Fixes incorrect permission check. Removes some unused stuff.
* MM-23832: Update permission to check.
* MM-23832: Switches to struct tags.
* MM-23832: Adds some docs for the permissions tag.
* MM-23832: Removes whitespace.
* MM-23832: Combines system admin restricted access with other acess-control tag.
* MM-23832: Fixes some tests.
* MM-23832: Clarifies docs, does not assume prior permission check in '-' access value case.
* MM-23832: Updates to correct access tag value.
* MM-23832: Adds test of the config settings tag access.
* MM-23832: Undoes whitespace change.
* MM-23832: Removes comment.
* MM-23832: Adds the permissions to the new roles rather than using OR conditions on the permissions checks.
* MM-23832: Removes or condition on permission check.
* MM-23832: Updates mapping.
* MM-23832: Typo fix.
* MM-23832: Adds new 'read_jobs' permission.
* MM-23832: Add read_jobs to all roles with manage_jobs.
* MM-23832: Adds new permission read_other_users_teams.
* MM-23832: Adds read filtering of config.
* MM-23932: Change tag value.
* MM-23832: Fixes some tests. Adds test for read config access tag.
* MM-23832: Adds permissions to list teams.
* MM-23832: Removes the '-' tag value. Adds a new permission read_channel_groups. Updates a permission check.
* MM-23832: Removes unnecessary parent permission for user_management. Fixes permission check change error.
* MM-23832: Removes unused parameter to filter/merge function.
* MM-23832: Renames migration name.
* MM-23832: Fix for godoc.
* MM-23832: Fixes tests.
* MM-23832: Only makes a map once rather than every function call. Doesn't require access tag on config field structs. Reverts one test update and fixes another.
* MM-23832: Removes all of the unnecessary uses of (*App).SessionHasPermissionToAny since removing the user_management parent permission.
* MM-23832: Updates constant type.
* MM-23832: Removes unnecessary comment.
* MM-23832: Renames permissions.
* MM-23832: Fix for permission name changes.
* MM-23832: Adds missing config access tags. Adds some requirec ancillary permissions for write_usermanagement_teams.
* MM-23832: Adds local API endpoint for getting config.
* MM-23832: If tag value is blank or restrict_sys_admin_write then don't do the permission check.
* MM-23832: nil check for strings prior to dereferencing.
* MM-23832: Fix for config display logic.
* MM-23832: Updates godoc.
* MM-23832: Delays the unrestricted check for parity with other permissions checks if the channel id does not exist.
* MM-23832: Removes tautology.
* MM-23832: Re-adds status code check.
* MM-23832: Adds new permission to edit brand image.
* MM-23832: Exports variable for use by mmctl.
* MM-23832: Initialize exported map for use by mmctl.
* MM-23832: Accept deprecated permissions as valid.
* MM-23832: Adds missing permissions to archive a channel.
* MM-23832: Adds missing permissions for managing team.
* MM-23832: Properly filters config values in patch and update API responses.
* MM-23832: Fixes license viewing and writing permissions.
* MM-23832: Require license to assign 'new system roles'.
* MM-23832: Adds translation keys.
* MM-23832: Updates translation order.
* MM-27529: Splits read_channel_groups into read_public_channel_groups and read_private_channel_groups.
* MM-23832: Prevent read-only permissions from editing site url test parameter.
* MM-23832: Prevent read permissions from sniffing ports and elastic password.
* MM-23832: Adds missing permission required for write user management channels.
* MM-23832: Allows new roles to search for channels.
* MM-23832: Adds ability for system_manager to manage jobs.
* MM-23832: Cluster status access by sysconsole permission, not manage_system.
* MM-23832: Adds 'add_user_to_team' permission to sysconsole write usermanagement teams.
* MM-23832: Fixes lint.
* MM-23832: Test fix.
* MM-23832: Test fix.
Co-authored-by: Catalin Tomai <catalin.tomai@mattermost.com>
Co-authored-by: Scott Bishel <scott.bishel@mattermost.com>
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
When we are reading from or putting to the cache, there is no need to deep copy
objects now because the cache is already serialized.
We also delete some lines from tests because the mock store directly returns the pointers
whereas actually they would be returned from the database where a serialization occurs again.
So we would be testing for the wrong thing by unnecessarily keeping the deep copies for DB reads.
https://mattermost.atlassian.net/browse/MM-27883
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* add warning as success result
* revert config changes
* add unit test, update unit test
* add a couple more tests
* update store layers
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* storetest/channel_store: fix flaky test for search too short term
* storetest/channel_store: use another stop word instead
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
Summary
- Attempt at fixing the AutocompleteInTeam/empty_string test which is flaky on MySQL-5.6. This is likely an index_merge_intersection optimization issue. See https://community.mattermost.com/core/pl/uzgygwa9e3d4pns14ayfs7mobw
for more discussion. The fix here simply modifies the where clause to change the query plan so index_merge_intersection is not used, while also not introducing a temporary table or a filesort.
Ticket Link
- https://mattermost.atlassian.net/browse/MM-26837
* MM-27040: Fix flaky tests due to same UpdateAt timestamp
Update queries are not guaranteed to always update the row.
Since we are only bumping the UpdateAt timestamps, if 2 update
requests hit concurrently within the same millisecond, then one
is bound to fail.
We fix all cases in the codebase where we were updating the UpdateAt
field to not check for count != 1, but rather count > 1.
A similar fix was already done in 3f46cf6f60.
* Remove incorrect test
Even after specifying all the 3 columns to coerce MySQL into choosing
the right multi-column index, it sometimes uses 2 separate indices
and does an index_merge of them.
This creates problems because the DeleteAt=0 search is essentially
the entire Posts table, and causes a disastrously bad performance
than even choosing the wrong index (idx_create_at).
The problem with this approach is that we hardcode the decision
to a specific index when MySQL was free to choose the right index
depending on table statistics. However, there does not appear to be
a case where this index can cause regressions than using some other index.
Another option here was to set optimizer_switch="index_merge_intersection=off"
at a session level for a transaction and then switch it back on after the
query is done.
However, this can cause some unintentional consequences because this setting
is only available at a session level and not at a query level.
There is no need to set something at the session level when an index hint suffices.
https://mattermost.atlassian.net/browse/MM-27575
* Refactor team store to use squirrel
* Fixing or expression
* Refactor team store to use squirrel
* Refactor team store to use squirrel
* Refactor team store to use squirrel
* Refactor team store to use squirrel
* Removing shadow declaration
* Fix error messages
* Improving Update() and Delete() calls
* Update store/sqlstore/team_store.go
Using squirrel to build 'in' clause
Co-authored-by: Miguel de la Cruz <mgdelacroix@gmail.com>
* Add missing translation
Co-authored-by: Dante Pippi <dantepippi@gmai.com>
Co-authored-by: Miguel de la Cruz <mgdelacroix@gmail.com>
Co-authored-by: mattermod <mattermod@users.noreply.github.com>
Co-authored-by: Claudio Costa <cstcld91@gmail.com>
