* rename strategy
* Update pkg/tsdb/sqleng/sql_engine.go
Co-authored-by: Torkel Ödegaard <torkel@grafana.org>
* more strict constraints
* Fixed so that it works on multi series results
* only apply the logic when original query returns 3 fields
* removed part of comment
* Update mysql test
Co-authored-by: Torkel Ödegaard <torkel@grafana.org>
Co-authored-by: Torkel Ödegaard <torkel@grafana.com>
Fixes panic/fatal error concurrent map writes in SQL data sources when multiple
queries are executed concurrently and you interpolate SQL query before executing it.
Fixes#35469
* draft pr convert opentsdb response to dataframes
* Add test for parse response and fix go lint
* Add test case for create request
* Use go-cmp in test
* Remove comment
Co-authored-by: Ida Furjesova <ida.furjesova@grafana.com>
* fix sql annotation parsing for empty responses
* fix backend when no data returned
* add back frontend changes
Co-authored-by: Ying WANG <ying.wang@grafana.com>
Move QueryData method into backend plugin manager which HandleRequest uses to
query data from plugin SDK supported data sources. This allowed us to remove a lot
of code no longer needed.
Ref #21510
Co-authored-by: Will Browne <wbrowne@users.noreply.github.com>
The /api/ds/query and /api/tsdb/query endpoints extract the intervalMs field from
each query in the request, but it currently seems to be ignored, at
least for queries to a Prometheus datasource. This changes
GetIntervalFrom function to check for the presence of intervalMs in the
query model if interval is missing.
* Migrate template variables straight into resource field,
Display error if unable to migrate
* Update workspace() template variable query function to return a resource URI
* update backend to support resource uris in workspace field
* don't load resource names for variables
* throw error when workspace migration returns no results
* update testDatasource to work with either resource URI or workspace GUID
* fix tests
* cleanup types just a little bit, as a treat
Uses new httpclient package from grafana-plugin-sdk-go introduced
via grafana/grafana-plugin-sdk-go#328.
Replaces the GetHTTPClient, GetTransport, GetTLSConfig methods defined
on DataSource model.
Longer-term the goal is to migrate core HTTP backend data sources to use the
SDK contracts and using httpclient.Provider for creating HTTP clients and such.
Co-authored-by: Arve Knudsen <arve.knudsen@gmail.com>
* wip
* wip:
* ui work for resource picker
* disable rows when others are selected
* resource picker open button
* Azure Monitor: Connect to backend with real data (#34024)
* Connect to backend with real data
* Fixes from code review.
* WIP:Begin thinking about how to make queries with scope determined by the resource picker
* More fixes post code review.
* Fixes after rebasing
* Remove outdated todo
* AzureMonitor: Support any resource for Logs queries (#33762)
* Apply button for resource picker
* scroll table body
* secondary cancel button
* loading state for nested rows
* Display resource components in picker button
* fix tests
* fix icons
* move route function
* Migrate from workspace to resource uri for log analytics (#34337)
* reword backwards compat comment
* remove base url suffix
* fix lint error
* move migrations to seperate file
* cleanup
* update regex
* cleanup
* update plugin routes to use new azure auth type
* use AzureDataSourceInstanceSettings alias
Co-authored-by: Sarah Zinger <sarahzinger@users.noreply.github.com>
* Bar chart label positioning and sizing
* Dev dashbard
* Improve autosizing
* Remove sync option
* Unify text sizing options between stat-ish visualizations and bar chart
* Add simple categorical data scenario and update dev dashboard
* Remove unused options builder
* Add docs annotations
* Fix go lint
* Determine Azure cloud based on authentication type
* Route resolution for clouds
* Fix tests
* Constants for cloud names
* move arg cloud route logic to routes.go
Co-authored-by: Erik Sundell <erik.sundell87@gmail.com>
* Add Azure Resource Graph in Azure Plugin
* fix lodash import
* Fix mock queries
* use "backend" sdk
* Address comments
* add converter for object type
* Query error should cause 400 & apply template var
* fix backend test & add documentation
* update image
* Address comments
* marshal body from map
* use interpolated query instead of raw query
* fix test
* filter out empty queries
* fix go linting problem
* use new query field language name
* improve variable tests
* add better tests for interpolate variable
Co-authored-by: joshhunt <josh@trtr.co>
Co-authored-by: Erik Sundell <erik.sundell87@gmail.com>
* add support for handling preprocessors in the backend
* add preprocessor tests
* use uppercase for constants
* add super label component
* remove error message from query editor since its not working (probably cause onDataError doesnt work anymore)
* use cheat sheet instead of help
* add return type annotation for projects
* add support for preprocessing. replace segment comp with select. change components names and refactoring
* cleanup
* more pr feedback
* fix annotations editor
* rename aggregation component
* fix broken test
* remove unnecessary cast
* fix strict errors
* fix more strict errors
* remove not used prop
* update docs
* use same inline label for annotation editor
* fix react prop warning
* disable preprocessing for distribution types
* using new default values for reducer
* auto select 'rate' if metric kind is not gauge
* fix create label format
* pr feedback
* more pr feedback
* update images
* add tests
* CloudWatch: Allow use of missing AWS namespaces using custom metrics
* CloudWatch: Allow use of missing AWS namespaces using custom metrics
* Devenv: add block for es 5.0, provisioned datasource & dashboard
* Trasnsform script property based on running ES version
* Handle different scripts format in BE
* add uninstall flow
* add install flow
* small cleanup
* smaller-footprint solution
* cleanup + make bp start auto
* fix interface contract
* improve naming
* accept version arg
* ensure use of shared logger
* make installer a field
* add plugin decommissioning
* add basic error checking
* fix api docs
* making initialization idempotent
* add mutex
* fix comment
* fix test
* add test for decommission
* improve existing test
* add more test coverage
* more tests
* change test func to use read lock
* refactoring + adding test asserts
* improve purging old install flow
* improve dupe checking
* change log name
* skip over dupe scanned
* make test assertion more flexible
* remove trailing line
* fix pointer receiver name
* update comment
* add context to API
* add config flag
* add base http api test + fix update functionality
* simplify existing check
* clean up test
* refactor tests based on feedback
* add single quotes to errs
* use gcmp in tests + fix logo issue
* make plugin list testing more flexible
* address feedback
* fix API test
* fix linter
* undo preallocate
* Update docs/sources/administration/configuration.md
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
* Update docs/sources/administration/configuration.md
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
* Update docs/sources/administration/configuration.md
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
* fix linting issue in test
* add docs placeholder
* update install notes
* Update docs/sources/plugins/marketplace.md
Co-authored-by: Marcus Olsson <marcus.olsson@hey.com>
* update access wording
* add more placeholder docs
* add link to more info
* PR feedback - improved errors, refactor, lock fix
* improve err details
* propagate plugin version errors
* don't autostart renderer
* add H1
* fix imports
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
Co-authored-by: Marcus Olsson <marcus.olsson@hey.com>
* Elasticsearch: use proper semver strings to identify ES version
* Update BE & tests
* refactor BE tests
* refactor isValidOption check
* update test
* Update pkg/tsdb/elasticsearch/client/client.go
Co-authored-by: Piotr Jamróz <pm.jamroz@gmail.com>
* Update pkg/tsdb/elasticsearch/client/search_request_test.go
Co-authored-by: Piotr Jamróz <pm.jamroz@gmail.com>
* Remove leftover FIXME comment
* Add new test cases for new version format
* Docs: add documentation about version dropdown
* Update docs/sources/datasources/elasticsearch.md
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
* Update docs/sources/datasources/elasticsearch.md
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
* Update docs/sources/datasources/elasticsearch.md
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
* Update provisioning documentation
Co-authored-by: Piotr Jamróz <pm.jamroz@gmail.com>
Co-authored-by: achatterjee-grafana <70489351+achatterjee-grafana@users.noreply.github.com>
Convert SQL data sources to return data frames.
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
Co-authored-by: Arve Knudsen <arve.knudsen@gmail.com>
Co-authored-by: Will Browne <will.browne@grafana.com>
Co-authored-by: Hugo Häggmark <hugo.haggmark@gmail.com>
* Elasticsearch: Add generic support for template variables
* format MovingAverage settings as numbers
* Move formatting logic to query builder & forma serial_diff settings as numbers
* modify presence check
* add todo
* minor fixes
* transform string values to numbers
* Move casting logic
* Slightly cleaner implementation
* Add BE tests
* Leverage elastic validation when string doesn't resolve to a numeric value
* move newly introduced test to testify
* add FE query_builder tests
* check error
* Parse values to float instead of int
* Fix tests & ParseFloat bit size
* Elasticsearch: Allow omitting field when metric supports inline script
* Add tests for MetricEditor to show a None option
* Add tests for useFields hook
* Alerting: allow elasticsearch metrics without field
For security reasons, log any SQL connection error
details rather than returning it to the client.
Fixes#26623Fixes#22000
Co-authored-by: Arve Knudsen <arve.knudsen@gmail.com>
This pull request migrates testdata to coreplugin streaming capabilities,
this is mostly a working concept of streaming plugins at the moment,
the work will continue in the following pull requests.
* first pass
* add instance manager
* fix tests
* remove dead code
* unexport fields
* cleanup
* remove ds instance from executor
* cleanup
* inline im
* remove old func
* get error working
* unexport field
* let fe do its magic
* fix channel name
* revert some tsdb changes
* fix annotations
* cleanup
* Return dataframe directly from the backend
* Streamline some transforms
* Fix lint issues
* Remove unused lib
* Fix datasource test
* Fix imports and add some typings
* Fix the typings and some tests
* Add private doc comment
* Remove private tag
* Add comments
* Fix some API docs issues
* SSE/InfluxDB: Change InfluxQL to work with server side expressions
updates the request to handle EpochMS as the time range (which is what expressions sends)
and the response to use millisecond timestamps instead of seconds.
fixes#31604
similar to https://github.com/grafana/grafana/pull/31455
* Convert Prometheus query response to dataframes
* Add parseResponse tests
* Fixed lint imports
* Force use UTC as timezone in timeVector
* Updated test to check timezone
* Tempo: Support opentelemetry response
* Tempo: convert Tempo to backend data source
* Update data source test
* Fix lint issues
* Apply suggestions from code review
Co-authored-by: Arve Knudsen <arve.knudsen@gmail.com>
* Return error when trace not found
Co-authored-by: Arve Knudsen <arve.knudsen@gmail.com>
* Stop using transformation to tables
* Update datasource.ts to parse from dataframes
* Add test on transformToDataframe func
* Remove parseResponse and do the work inside parseToAnnotations method
* Name the dataframe after the RefID
* Small fix when mapping to annotation
* Add tests in annotation_query_test.go class
* Small fix in loop iteration
* Change for loop signature
Changes the time value of the response to be milliseconds, which I think is like the other data sources I have seen.
Also if it sees the timerange values as numbers, assumes they are unixMS (and come from expressions, or our new backend.TimeRange type). When this timerange is used, the request is sent to graphite with from and until as epoch timestamps
* Add configuration in ConfigEditor and default to 1000
* Show data in explore if any even if there is an error
* Update pkg/tsdb/influxdb/flux/executor.go
* Better handling of defaults
* Add test for runQuery to show data even with error
* Update public/app/store/configureStore.ts
Co-authored-by: Giordano Ricci <gio.ricci@grafana.com>
* Update public/app/plugins/datasource/influxdb/components/ConfigEditor.tsx
Co-authored-by: Ivana Huckova <30407135+ivanahuckova@users.noreply.github.com>
* Update tooltip
* Update input
* Lint fixes
* Update snapshots
* Update decorator tests
Co-authored-by: Giordano Ricci <gio.ricci@grafana.com>
Co-authored-by: Ivana Huckova <30407135+ivanahuckova@users.noreply.github.com>
* update interval to fix no unit
* fix bug
* update with new method
* format go
* add test
* change test method to add test error nil
* add simplejson and model package
* change json format
* add parentheses
* add simplejson to function
* change mode to model
* move assert import package
* add one enter to avoid goimport error
* change to test case
Converts the core testdata data source to use the SDK contracts and by that
implementing a backend plugin in core Grafana in similar manner as an external one.
Co-authored-by: Will Browne <will.browne@grafana.com>
Co-authored-by: Marcus Efraimsson <marefr@users.noreply.github.com>
Co-authored-by: Ryan McKinley <ryantxu@gmail.com>
* Add GraphView component
* Add service map panel
* Add more metadata visuals
* Add context menu on click
* Add context menu for services
* Fix service map in dashboard
* Add field proxy in explore linkSupplier
* Refactor the link creation
* Remove test file
* Fix scale change when view is panned
* Fix node centering
* Don't show context menu if no links
* Fix service map containers
* Add collapsible around the service map
* Fix stats computation
* Remove debug log
* Fix time stats
* Allow string timestamp
* Make panning bounded
* Add zooming by mouse wheel
* Clean up the colors
* Fix stats for single trace graph
* Don't show debug config
* Add more complex layout
* Update layout with better fixing of the root nodes
* Code cleanup
* Change how we pass in link creation function and some more cleanup
* Refactor the panel section into separate render methods
* Make the edge hover more readable
* Move stats computation to data source
* Put edge labels to front
* Simplify layout for better multi graph layout
* Update for dark theme
* Move function to utils
* Visual improvements
* Improve context menu detail
* Allow custom details
* Rename to NodeGraph
* Remove unused dependencies
* Use named color palette and add some fallbacks for missing data
* Add test data scenario
* Rename plugin
* Switch scroll zoom direction to align with google maps
* Do some perf optimisations and rise the node limit
* Update alert styling
* Rename function
* Add tests
* Add more tests
* Change data frame column mapping to use column names
* Fix test
* Fix type errors
* Don't show context menu without links
* Add beta status to panel
* Fix tests
* Changed function to standard methods
* Fix typing
* Clean up yarn.lock
* Add some UI improvements
- better styling of the zoom buttons
- disable buttons when max reached
* Fix panel references after rename
* Add panel icon
* cloud monitoring mql support
* reduce nesting
* remove resource type from deep link since. its removed for two reasons. first of all it is not needed for the link to work. secondly, by adding the resource type, the the link will differ from the query in grafana which I think is misleading
* use frame.meta.executedQueryString instead of legacy meta
Co-authored-by: Erik Sundell <erik.sundell87@gmail.com>
Adds support to the terms aggregation for ordering by percentiles and extended stats.
Closes#5148
Co-authored-by: Giordano Ricci <grdnricci@gmail.com>
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
* Convert Cloud Monitoring (Stackdriver) Datasource to use Dataframes #29830
* add deeplink into config
* omggggggggggggggg this deeplink works!
* move unit to the backend part
* remove unit from frontend
* only set the config fields[1] for deeplink and unit
* refactory + fix some test
* remove frontend test for unit
* adding backend test for unit mapping
* resolve review
* rewrtie unit logic to do exactly the same as frontend filter
* refactory
The current backend code doesn't honor the minimum interval set in the UI for alerts using
the Elasticsearch data source. This means that the data the alerts are triggering against
will never match the data in the visualization if auto is used in the date histogram as interval.
This fixes the problem to make sure that date histogram auto interval is set according to
min interval set in UI for the query or fallback to data source min interval setting.
Fixes#22082
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
* get meta data from dataframe meta
* gmdMeta is not used anymore
* wip: move deep link creation to backend
* Refactor backend. Remove not used front end code
* Unit test deep links. Remove not used frontend tests
* remove reference to gmdmeta
* fix goimports error
* fixed after pr feedback
* more changes according to pr feedback
* fix lint issue
* fix bad math expression check and fix bad test
* Decrease nesting
* put link on first data field only
* Elasticsearch: Add support for serial diff pipeline aggregation
* Removing settings transsforms
* Removing unused deps
* removing unused dep
* Fixing type in test
* Adding backend support for serial_diff
