* update vega version
* a11y skip test
* a11y skip test attempt 2
* adding back all a11y tests
* fix failed karma test
* remove extra 0BSD license
* coalesce yarn.lock versions a little
* update kbn/pm dist
* fix CI
* fix Vega View for ML
Co-authored-by: Michail Yasonik <michail.yasonik@elastic.co>
Co-authored-by: spalger <spalger@users.noreply.github.com>
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
This PR adds an API to the AlertsClient which allows users to execute actions immediately, rather than via TaskManager.
This also moves the TaskManager based execution to a new api on ActionsClient along side the immediate execution.
* Add copy agent config endpoint, schema types, and request hook
* Add copy config to context menu and modal
* Change data sources to be copied via bulk instead of singles in parallel
* Add copy endpoint to openapi spec file
* Add api integration tests
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* Don't use filter to get default config & output
the KQL parsing takes up a lot of CPU time. Avoid it with a search string
Also add a `refresh: false` as suggested by @bkobel
* Does restoring await fix CI?
* Remove 'refresh: false' from agent enroll
Does this this fix the failing test? It does keep the PR more focused
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* set wrap to the errorLink so it doesnt go outside of box
* updating
* set wrap to the errorLink so it doesnt go outside of box
* updating
* Show full message on hover
Co-authored-by: Nathan L Smith <smith@nlsmith.com>
[SECURITYSOLUTION][INGEST] Fixes endpoint link, adds link from endpoint details to fleet
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* [APM] Changes to duration formatting
* Change the threshold for showing microseconds to 1 millisecond instead of 10. This means you now get "900 µs/1.3 ms/20.0 ms" instead of "900 µs/1300 µs/20 ms."
* Change milliseconds formatted with `asDuration` to be `asDecimal` instead of `asInteger`. That means you get "0.0 ms/2.5 ms/3.0 ms" instead of "0 ms/2 ms/3 ms."
* Tables were all using their own module-scoped functions that called `asDuration` to format things as decimal milliseconds. Extract this to an `asMillisecondDuration` function exported from the duration helpers and use it in all the tables.
* Change `getResponseTimeseries` in chart selectors to use `asDuration` to make all chart timeseries units consistent.
* Don't export `convertTo` from the duration helpers as it's now not used anywhere. Always use a more specific exported function for more consistency.
* Change ">=" to "≥" in the ML flyout text.
* Update e2e snapshot
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* Ensure we use existing Elasticsearch config
* Use separate type for this to ensure custom properties work
* PR suggestions
* PR feedback
* PR feedback
* Fix type issues
* PR feedback
Added example for using dashboard container by value
1.1 Refactored embeddable explorer e2e test to use new example, removed not needed kbn_tp_embeddable_explorer plugin.
For embeddable explorer examples went away from using getFactoryById() to improve type checks
There is new component a replacement for EmbeddableFactoryRenderer with slightly more flexible api: EmbeddableRenderer.
3.1 We can improve it going forward to support more use case
* rename server plugin to siem to avoid privilege issue
* review from alerting
* missing change with rename
* fix tests
* missing api integration test
* fix api integration spaces