* Choose mode. Add log field on top level dualwriter
* Add logs
* Turn DualWriter into a full interface. Fix tests
* Lint
* Use struct for dualWriter interface
* Use struct
* Default should be legacyStore for all entities
* Fix test. Get rid of extra concrete type
* Remove comment
* Add comment
* Temp set dualwriter mode 2 for playlists while configs are not in place
* Add modes type + add comment on what each mode does
* Don't require watcher interface for now
* Use storage implementation on mode 2
* Update pkg/apiserver/rest/dualwriter_mode2.go
Co-authored-by: Todd Treece <360020+toddtreece@users.noreply.github.com>
* Pass log values to the context
* test
* Update pkg/apiserver/rest/dualwriter_mode3.go
Co-authored-by: Dan Cech <dcech@grafana.com>
---------
Co-authored-by: Todd Treece <360020+toddtreece@users.noreply.github.com>
Co-authored-by: Dan Cech <dcech@grafana.com>
* clean up error handling in postDashboard and remove UserDisplayDTO
* replace GetUserUID with GetUID and GetNamespacedUID, enforce namespace constant type
* lint fix
* lint fix
* more lint fixes
* Only let nodeGraph frames process through nodeGraph
* add metadata to test frames to be processed
* Revert "add metadata to test frames to be processed"
This reverts commit fc366a0ab4.
* Revert "Only let nodeGraph frames process through nodeGraph"
This reverts commit b8be50e219.
* Only send node graph frames to node graph logic
* remove unnecessary logic
* Loki: Add label values API selector to setting and use label_values API when selected
* remove CHANGELOG change
* add docs
* Support Loki 2.7+ only, so unconditionally use /labels API
* Correct doc for fetchLabels and fetchLabelValues functions
* Fixes after merge
* Do not encode query parameter twice
* return getLabelKeys in Completion Provider
* docs
* Add test for LabelParamEditor
* Update public/app/plugins/datasource/loki/LogContextProvider.test.ts
Co-authored-by: Ivana Huckova <30407135+ivanahuckova@users.noreply.github.com>
* toHaveBeenCalledWith
---------
Co-authored-by: Matias Chomicki <matyax@gmail.com>
Co-authored-by: Ivana Huckova <30407135+ivanahuckova@users.noreply.github.com>
* added admonition and API info
* Update docs/sources/administration/roles-and-permissions/access-control/_index.md
---------
Co-authored-by: Misi <mgyongyosi@users.noreply.github.com>
Co-authored-by: Jack Baldry <jack.baldry@grafana.com>
* Add OrgRoleMapper
* Address feedback, add more tests
* Prevent resetting assignments when global org mapping fails
* Provide a parsing function instead of recalculating the org_mapping on every mapping
* Introduce strict role parsing/mapping
* Introduce MappingConfiguration
* Handle other edge case
* Add tests
* lint
* Apply documentation update suggestions
* Apply GetDefaultOrgMapping suggestions from code review
* Apply suggestions for cleaning up unnecessary err in ParseOrgMappingSettings
* Apply suggestions from code review
* Address feedback suggestions
* Cleanup
* Reduce cognitive complexity of ParseOrgMappingSettings
---------
Co-authored-by: Gabriel MABILLE <gamab@users.noreply.github.com>
* Get custom cloud list from grafana-azure-sdk-go for frontend settings
* Remove getAzureClouds() and call CustomClouds() directly
* remove unused type
* fix: reloaded scene first step topScene
The top scene was not being initialized
when reloading trails, causing it to use
the top scene of whatever was selected
previously.
* fix: select new metric node not reloading right
On reloading trails, any metric node that
was 'green' with a undefined metric due to
clicking on the "select new metric" button,
was reloading with the metric of the step
directly before it (in index order).
* fix: opt-in to `useQueriesAsFilterForOptions`
With updating the scenes library, dashboards must now explicitly
opt in to `useQueriesAsFilterForOptions`.
* fix: loaded trails modified parent var on new step
This ensures that recently loaded trails won't have variable changes
which create new steps modify the previous step.
* Update Yarn to v4.2.1
* add new yarn version
---------
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
Co-authored-by: Ashley Harrison <ashley.harrison@grafana.com>
Storage Server: avoid incorrect parsing of connection string
In storage server, a connection string is generated to connect to the
backing database. In case of Postgres, we were generating the string
like `"user=grafana pass= host=postgres.example port=4567"`.
This triggered an edge case in `pq` (the go postgres driver) to parse
`pass` to be equal to `host=postgres.example`, and host being reset to
default value.
Using single quotes in the connection string fixed this.
Signed-off-by: Prem Kumar <prem.saraswat@grafana.com>
Add reqAction to Prometheus routes
This should allow provisioned service accounts that do not have a role to use the /proxy endpoints for Prometheus. This
is desired for some integrations that need to query for labels or other data not part of /ds/query.
