* convert it to functional component
* useReducer
* usePromQueryFieldEffects
* clean up the code
* remove localStorage provider
* introduce usePromQueryFieldEffects.test.ts
* simpler state management
* remove mocks
* linting + betterer
* Split PrometheusMetricsBrowser into small functional components
* remove /series endpoint usages for Metrics Browser
* add time ranges
* remove unused file
* fix unit test
* use limit
* re-fetch labels when a label value is selected
* reset the label keys when selector is empty
* select multiple labelValue
* fix unit tests
* add comments
* don't show unpresented labelKeys
* don't clear out selected label keys that are in local storage
* show values properly
* fix clearing
* don't discard already selected label values when a new label value selected
* remove unnecessary unit test
* clear all selections
* if metrics list is empty remove all label keys too
* re-populate values when all label selections are removed
* safety
* more comments
* updating the whole fetching mechanism
* buildSafeSelector
* loadSelectedLabelsFromStorage
* fetchMetrics
* fetchLabelKeys
* fetchLabelValues
* more unit test and comment
* timeRange changes will trigger re-fetch
* some more comments
* fix the unit test
* lint
* fix unit tests
* fix timerange causing problem
* remove fethcing message
* proper initialization
* fix value fetching during metric selection
* fix value fetching
* fix unit tests
* fix a bug
* comments
* some precaution
* fix the test
* fetch metrics when metric selection is changed
* trigger initialize when seriesLimit is change
* Revert "trigger initialize when seriesLimit is change"
This reverts commit c6cbbc1386.
* remove redundant comments
* e2e: improves flakiness
* Chore: refactor flaky test
* e2e: more refactor
* e2e: do not keep successfull test run videos
* e2e: do not log selectors
* chore: updates after pr feedback
* chore: updates after pr feedback
* chore: adds retries to flaky e2e test
* e2e: skip flaky tests
* e2e: skip flaky tests
* e2e: revert back to timeout
* chore: removes all the skips
Throw errors when trying to call List in legacy with an RV specified
Change ResourceVersion to microseconds
Co-authored-by: Marco de Abreu <18629099+marcoabreu@users.noreply.github.com>
* Alerting: Update docs for RBAC
This updates the documentation for RBAC to match the changes from 032299011a
* add note about RBAC permissions with datasource permissions
* writers toolkit update
* Rename and make folder migrator public
* Move skipRepo out of tree
* Move folder logic out of worker
* Use wrapper function for clone
* Introduce storage swapper
* Use RepositoryResources interface
* Add TODO
* Signature with option
* Use plural for folders
* Split more logic out of worker
* Refactor further
* Remove todo
* Refactor further
* Test worker
* Add test for unified storage
* Add mock for bulk store
* Clean line
* Test more about storage
* Happy path storage test
* Finish storage tests
* Add more interfaces
* Add some tests legacy
* Complete test coverage legacy
* Complete coverage legacy
* Fix compilation
* Add initial folders tests
* Finish coverage for folders
* Test namespace cleaner
* Fix typo
* Add some tests for resources
* Finish coverage migrate package
* Update tests
* Fix jobs tests
* Unit test users
* Convert user tests to table tests
* Fix linting in tests
* Fix typo
* Use PreloadAllUserInfo
* Add FIXME
* SQL Expressions: Bump GMS for STDDEV(), VARIANCE()
Bump Go MySQL Server to the latest version, to get support for STDDEV()
and VARIANCE() functions, and their aliases.
See https://github.com/dolthub/go-mysql-server/pull/2928
* Update app-code to comply with latest GMS version
* Run `go mod tidy`
* Update test which had depended on STDDEV being missing
* Run `make update-workspace`
Alerting: fix flaky tests
Some test conditions introduced in #103403 are flaky because they rely on random behavior of the generator.
Sometimes rules are generated with an updated by (which warrants the lookup of the users). This makes it so those tests which are checking the user lookup always have rules with updated by.
What is this feature?
Send resolved notifications not only when an alert state becomes stale (series is missing) and transitions from Alerting to Normal, but also from Error, NoData and Recovering.
Why do we need this feature?
Previously, when an alert state became stale or was deleted, it would transition to Normal but wouldn't trigger resolved notifications to the Alertmanager. This meant we relied on the Alertmanager to send resolved notifications when the alert expires. However, if the Alertmanager state is lost, these resolved notifications would never be sent, leaving users with firing alerts in their notification channels. This PR ensures that any transition from a firing state (Alerting, Error, NoData, Recovering) to Normal triggers a resolved notification.
Laura Fernández
Artur Wierzbicki
Serge Zaitsev
ismail simsek
Hugo Häggmark
Marco de AbreuandMarco de Abreu
grafana-pr-automation[bot]andgithub-actions[bot]
Gilles De MeyandMatthew Jacobson
wuwengandDave Henderson
Nick Richmond
Moustafa Baiou
Matias ChomickiandSven Grossmann
Kevin Minehart
Edvard Falkskär
Roberto Jiménez Sánchez
Andres Martinez Gotor
Sonia Aguilar
Tom Ratcliffe
Gábor Farkas
Esteban Beltran
Sam Jewell
Alexander Zobnin
Ashley Harrison
Andrej Ocenas
Torkel Ödegaard
renovate[bot]
Alex Khomenko
Juan Cabanas
colin-stuart
Ryan McKinley
Alexander Akhmetov
Kristina
Georges Chaudy
Ida Štambuk