-
Notifications
You must be signed in to change notification settings - Fork 8.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[savedObjects] use index template rather than healthcheck #14142
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The SavedObjects code defines `isEsUsable$` and `testEsAvailability$` to negotiate whether SavedObjectClients should be attempting to communicate to elasticsearch. - `isEsUsable$` - can be either `undefined`, `true`, or `false` - if `isEsUsable === undefined` - All requests are paused - if `isEsUsable === false` - Send 503 to all paused and future requests - if `isEsUsable === true` - Request are sent to elasticsearch - `testEsAvailability$` - emits when: - elasticsearch plugin is accessible - a request to es is blocked - es does not respond, or responds with a status >= 500 - on each event: - set `isEsUsable` to `undefined` - attempt to put the index template to elasticsearch - set `isEsUsable` to wether putTemplate was successful
…nd something else
spalger
changed the title
[SavedObjects] use index template rather than healthcheck
[savedObjects] use index template rather than healthcheck
Sep 23, 2017
Merged
This was referenced Sep 23, 2017
I'm putting this on ice, using #14202 solves my immediate need |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
WIP
Based on #14141
Fixes #12819
Completely removes SavedObject index management from the health check and to own internal "esAvailability" check.
I’m working on a Kibana feature that will see tens, if not hundreds, of config documents created. Since that is currently the job of the health check and it’s impractical to iterate and check for every expected config document every 2.5 second, we need a more reactive approach; UiSettings should just automatically create the config doc when it doesn’t exist. Unfortunately, we have grown to rely on the health check so much that we often just assume that the Kibana index will always be usable, and we can just write to it without checking, but that really isn’t the case. Writing to the Kibana index without knowing for sure that the index exists is dangerous (see #12819) and the solution recommended by the Elasticearch team is use index templates.
To do this we get rid of the index creation in the healtch check, wrap the callCluster function passed to the SavedObjectCient so that every request to elasticsearch is predicated by a “check”, rather than rely on an interval. This check currently attempts to write an index template to the kibana index (throttled to one attempt per second) when:
callCluster
rejects with a status code >= 500callCluster
rejects without a status codeSavedObjectsClient
uses itscallCluster
function after a failed checkWe still need to patch the Kibana index if it already exists and we need a 🐃 of tests, but I'm excited by how this is coming along.