This could cause scripts to fail when kafka is not available even if not
used.
KafkaProducer class is already a singleton wrapper so this changes no
behavior
Latest migrations have been applied.
Only thing we need to keep in sync is CLICKHOUSE_MIGRATION_HOSTS env
variable going forward - managed via secrets manager.
* Use `statshog` over python-statsd
More support for tags!
* Include custom tags for every query + add annotation to query
After this we can:
- Figure out from query logs where queries are coming from (speeding up debugging)
- Break down query speeds by user queries vs others (e.g. celery) --
better represents overall speed
- Can figure out how fast queries are on average for various teams
* Use tags in more queries over interpolation
This way we can set up more interesting graphs \o/
* Solve mypy error
* Fix a flaky test (due to ordering)
* remove wrap on plugin logs table
* Implement a real fix by checking if ResizableTableProps.expandable is set
Co-authored-by: Michael Matloka <dev@twixes.com>
* Refactor plugin installation and reloading logic
* Preinstall and autoenable plugins (only GeoIP one for now)
* Fix PluginManager.install issues
* Add tests
* Disable plugin preinstallation on Cloud
* Check plugins access level before preinstallation
* Rename field `preinstalled` to `is_preinstalled`
* Adjust tests for Cloud v. self-hosted
* Fix test_preinstalled_are_autoenabled
* Ensure install plugins access level for preinstallation tests
* Bump migration
* Bump migration
* Add and use Plugin.get_default_config, with tests
* Refactor private PluginConfigSerializer methods out of the class
* Skip preinstallation of plugin if it couldn't be done (e.g. airgapped)
* Don't hit GitHub API in tests
* Satisfy mypy
* Fix a test on Cloud CI
* Revert "Revert "Increase Element model varchar limits (#3912)" (#3923)"
This reverts commit e30681bade.
* Use RunSQL to avoid ORM crap
* Fix attr_class being converted from varchar[] to varchar
* Bump migration
* Simplify migration and remove attr_class alter due to locking
* Bump migration
We're seeing queries try to spawn 15k threads due to an invalid
max_threads setting in clickhouse. Since no-one awake has ssh access
this is a temporary hotfix to make things work.
