* Added coverage around search and dashboard tests.
* Added tests to check for whether the resource is available. If not, return 404.
* Skipped two tests due to https://github.com/elastic/kibana/issues/19713. Added error handling for relationships API for when no result is found. Return 404.
* Applied patch file per PR.
* Applied Chris patch and tested locally. No failures.
* Removed ajv and utilised joi for schema validation.
* Fixed package.json.
* Copied package.json description from master.
* Reverted package.json and made proper edit.
Fixes#20017
Currently the xpack_main plugin only sets its status when the elasticsearch plugin changes status, which is a fine signal that the plugin should recheck for the license info, but not the only time it can be updated. At some point recently a race condition became more likely to be triggered, which is detailed in https://github.com/elastic/kibana/issues/20017#issuecomment-398209354 and causes the xpack_main plugin to set its status to red and never return it to green, even after the license info is successfully fetched from Elasticsearch.
To fix this the `XpackInfo` class now exposes the `onLicenseInfoChange(handler)` method which will call the handler whenever the xpack info changes. This way the xpack_main plugin can reflect the current state of the license info in its status and doesn't have to wait for the elasticsearch plugin to change status again.
To test this I recommend running kibana with `--elasticsearch.healthCheck.delay=500` or even `250`, and then starting Elasticsearch after Kibana has started. Without this change Kibana is very likely to attempt a healthcheck between the time that Elasticsearch is "started" and the x-pack license info is completely loaded, triggering the race condition. With this change you should see the xpack info plugin (and all other xpack plugins) go red with a more informative message, and then after the first request or 30 second polling timer, the plugins should go green again.
* Add Inspector feature
* So long, and thanks for all the fish, spy panel
* Fix several functional tests
* Fix unit tests
* Fix spy panel button tests
* Replace old spy panel documentation
* Disable test temporarily until we have dashboard triggers
* Enter edit mode for dark theme test
* Fix some more functional tests
* Fix more functional tests
* More test fixing
* Fix more functional tests
* Allow opening the inspector via loader handler
* Refactor InspectorViewChooser, remove unused CSS
* Remove dead code
* Fix data download button style
* Remove redundant code
* Load inspectorViews for dashboard_viewer
* Extract inspector views to custom core_plugin
* Switch API to TypeScript 🎉
* Design changes
* Remove icons from views
* Design changes
* Improve typings of API
* Add typing to all adapters
* Show loading spinner in request selector
* Rewrite InspectorView to TypeScript
* Fix help text for data view
* Remove deprecated React lifecycle methods
* Embed inspector into dashboard panel actions
* Remove temporary inspector trigger
* Remove old CSS
* Fix dashboard trigger for new panel action
* Add tests for InspectorPanel and DataAdapter
* Produce a hierarchical table if the vis is hierarchical
* Remove allowJs option again
* Add missing Apache license headers
* Close inspector on dashboard when navigating away
* Use proper title for dashboard panels
* Fix functional tests
* Skip broken test for now
* Flush view chooser button
* Add request adapter tests
* Skip more tests, broken due to typescript
* Add Request Time description
* Add description for courier request
* Fix tests
* Replace icon by new (not yet released) icon
* Finalize design of inspector
* Remove discover test, that relied on spy panels
* Change API to be properly mockable in tests
* Add aria-live region for request status
* Replace old method in functional tests
* Replace abitrary magic number
* Use object destructuring in vis
* Fix issue with crashing requests view
* Add request time tooltip
* Get request body of correct search source
* Make filter buttons properly keyboard accessible
* Follow Dave's design suggestions
* Remove redundant request from name
* Remove unneeded comments
* WIP raw-formatted values
* Fix filtering issue
* Fix tests and more license headers
* Add data view tests
* Remove search from table
* Fix typos
* Implement review suggestion
* Remove artificial delays for testing
* Fix new panel action structure
* Minor design adjustments
* Fix failing functional test
* Update failing snapshot test
* Implement final wording
* Apply new EUI styling
* Fix closing inspector in tests
* Fix sorting of table
* Align punctuation between tooltips
I've seen a few failures recently like https://kibana-ci.elastic.co/job/elastic+kibana+pull-request+multijob-selenium/4940/console which are caused by the modal not being visible, but if you look at the screenshot you can see the the modal was just about to be displayed but the find didn't wait quite long enough. ~~To fix this temporarily I've increased the `isConfirmModalOpen()` timeout from 2 to 5 seconds.~~
These tests are actually trying to assert that the modal is gone, so rather than simply checking the state of the modal once, the `CommonPage#isConfirmModalOpen()` method has been renamed to `CommonPage#expectConfirmModalOpenState()` which expects a boolean and will assert the state of the confirm modal within a retry, so that as long as the modal eventually transitions to that state it will eventually resolve.
* Patch ConfigView changes from x-pack-kibana to OSS Kibana.
* Remove old css.
* Update style for queue statement.
* WIP modifying based on designer feedback.
* Add flatten function and list class.
* Rename functions to be more descriptive.
* WIP checkin. Modify components to handle flattened object list.
* Finish moving flatten logic into classes, add tests.
* Simplify flattening, remove non-native dependency. Add more tests.
* Add defaults to simplify function call.
* Refactor two blocks into a function.
* Begin adapting components for list.
* Add collapse functionality.
* Add expand functionality.
* Nested collapsed elements remain collapsed on parent expansion.
* Style section headers.
* Update Plugin statement styles.
* Add DetailDrawer support.
* Update statement formatting.
* Add stats to plugin element rows.
* Resolve conflicts.
* Remove obsolete code.
* Reorganize code.
* Remove warnings.
* Add PropTypes. Add keys to arrays and iterables.
* Update color for borders/buttons.
* Add stat class. Clean up code.
* Convert plugin statement component from class to function.
* Update queue metrics message.
* Update style to make view more responsive.
* Change section heading size.
* Remove gutter from metrics group.
* Change name "stat" to "metric".
* Remove obsolete export line, simplify declaration, based on PR feedback.
* Add new functional component in place of model class.
* Add PropTypes to several components. Rename a function. Add keys to metrics.
* Convert stateless classes to functional components.
* Prefer ES6 syntax over bindings for Component methods.
* Do not render statement section if there are no statements.
* design cleanup for pipeline viewer
* Update CSS to add min-height to page.
* Rename "elements" to "statements". Delete unused LESS variables.
* Revert naming of "statements" to "elements" for StatementList component.
* Update jest snapshots for DetailDrawer.
There are several cases where we want to do different things in the
optimizer based on whether we are running in the distributable version
of Kibana or running from source. Historically we have handled this by
relying on the env, but as we start to prebuild things, like TypeScript
and soon SCSS, we actually need to know if the version of Kibana running
was a result of the build process or not. To address this the src/utils
module now exports a constant: `IS_KIBANA_DISTRIBUTABLE`. This constant
will be based on the build.distributable flag being set in Kibana's
package.json, which is done automatically in the build process.
This PR upgrades RxJS to version 6 and switches to a fork of `stream-to-observable` which includes an updated version of `any-observable` that supports RxJS 6 until https://github.com/jamestalmage/stream-to-observable/pull/10 is merged. The primary change in this version of RxJS is the movement of stream operators from `Obersable.prototype` to the `rxjs/operators` module. Some of the operators, like `catch` and `do`, have been renamed (`catchError`, and `tap`). The Obsevable factories have also been moved from static methods on the `Observable` class to named exports of the root `rxjs` module. Some of those factories have also changed slightly, like `fromEvent` which now emits arrays if the event handler is called with multiple arguments.
```js
// import the Rx namespace to get the Observable factories
import * as Rx from 'rxjs';
// import the operators as named imports
import { map, tap, switchMap } from 'rxjs/operators';
```
* Migrate dashboard save error from old toast message to new EUI toast
* save function does not need to exist on anymore
* account for error in onSave
* update clone
* lint errors
This PR tries to reduce the Elasticsearch cluster load for index patterns with many indices.
**Theory of operation**
Before this PR, Elasticsearch had to perform the query and sorting for every shard matching the index pattern. In order to avoid that in a time-base indexing scheme, the queries should include a `range` filter. This enables Elasticsearch to rewrite most of the shard accesses to `match_none`. But since the context view operates on document counts, the time intervals need to be determined heuristically:
* start of the interval is the second end of the previous query iteration or `anchor_time` if it is the first iteration
* end of the interval is `interval_start +/- n days` with `n in {1, 7, 30, 365, 10000}` or unlimited if insufficient hits were returned for all `n`
This date arithmetic introduces the assumption that the primary sorting field is a date or at least numeric. Therefore, the `sortingField` has been renamed to `timeField` to make those new assumptions explicit.
**Other notes**
As an additional optimization, the queries are now executed in a `constant_score` filter context to enable caching by Elasticsearch.
Tests for `fetchSuccessors` and `fetchPredecessors` were added.
Changes in ElasticSearch required a concurrent fix of #17696, which is also included. It now splits up the anchor `uid` into `anchorType` and `anchorId` and uses them in an `ids` query.
**Testing**
The fact that only a small subset of the shards are involved in a query should be observable using the `skipped` shard count in the response.
fixes#15143fixes#17696
* [Monitoring/No Data] Show a "still waiting" link after 19s
* minor bug fix: spinner if _interval_ updating
* update "we tried"
* some wording changes on activating / trying to activate monitoring
* cancel the setTimeout when component gets unmounted
* more wording tweaks
* update snapshot
