[ML] Clean up inference indices on failed endpoint creation #136577
Merged
Conversation
This file contains hidden or 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
Prior to this change, if only one of the .inference or .secrets-inference indices was updated when creating an inference endpoint, the endpoint creation would fail, but the successfully written doc was not removed, leading to inconsistent document counts between the two indices. This commit removes any documents that were written in the case that a partial failure occurred, but does not change the behaviour in the case where no updates to the indices were made. - Invoke a cleanup listener if a partial failure occurred when storing inference endpoint information in the .inference and .secrets-inference indices - Refactor ModelRegistry to use BulkRequestBuilder.add(IndexRequestBuilder) instead of the deprecated BulkRequestBuilder.add(IndexRequest) - Include cause when logging bulk failure during inference endpoint creation - Add integration tests for the new behaviour Closes elastic#123726
DonalEvans
added
>bug
:ml
|
Pinging @elastic/ml-core (Team:ML) |
|
Hi @DonalEvans, I've created a changelog YAML for you. |
jonathan-buttner
approved these changes
Oct 15, 2025
| } | ||
|
|
||
| logBulkFailures(model.getConfigurations().getInferenceEntityId(), bulkItemResponses); | ||
| boolean anySuccess = false; |
There was a problem hiding this comment.
nit: It might be slightly cleaner to move this boolean above the cleanupListener. Then in cleanupListener we can have a check for anySuccess, if there was a success, then do the delete.
Then we don't need to track listenerToInvoke.
DonalEvans
added
auto-backport
This was referenced Oct 16, 2025
DonalEvans
added a commit
to DonalEvans/elasticsearch
that referenced
this pull request
Oct 16, 2025
…136577) Prior to this change, if only one of the .inference or .secrets-inference indices was updated when creating an inference endpoint, the endpoint creation would fail, but the successfully written doc was not removed, leading to inconsistent document counts between the two indices. This commit removes any documents that were written in the case that a partial failure occurred, but does not change the behaviour in the case where no updates to the indices were made. - Invoke a cleanup listener if a partial failure occurred when storing inference endpoint information in the .inference and .secrets-inference indices - Refactor ModelRegistry to use BulkRequestBuilder.add(IndexRequestBuilder) instead of the deprecated BulkRequestBuilder.add(IndexRequest) - Include cause when logging bulk failure during inference endpoint creation - Add integration tests for the new behaviour - Update docs/changelog/136577.yaml Closes elastic#123726 (cherry picked from commit 9abc0bd)
💚 All backports created successfully
Questions ?Please refer to the Backport tool documentation |
DonalEvans
added a commit
to DonalEvans/elasticsearch
that referenced
this pull request
Oct 16, 2025
…136577) Prior to this change, if only one of the .inference or .secrets-inference indices was updated when creating an inference endpoint, the endpoint creation would fail, but the successfully written doc was not removed, leading to inconsistent document counts between the two indices. This commit removes any documents that were written in the case that a partial failure occurred, but does not change the behaviour in the case where no updates to the indices were made. - Invoke a cleanup listener if a partial failure occurred when storing inference endpoint information in the .inference and .secrets-inference indices - Refactor ModelRegistry to use BulkRequestBuilder.add(IndexRequestBuilder) instead of the deprecated BulkRequestBuilder.add(IndexRequest) - Include cause when logging bulk failure during inference endpoint creation - Add integration tests for the new behaviour - Update docs/changelog/136577.yaml Closes elastic#123726 (cherry picked from commit 9abc0bd)
Kubik42
pushed a commit
to Kubik42/elasticsearch
that referenced
this pull request
Oct 16, 2025
…136577) Prior to this change, if only one of the .inference or .secrets-inference indices was updated when creating an inference endpoint, the endpoint creation would fail, but the successfully written doc was not removed, leading to inconsistent document counts between the two indices. This commit removes any documents that were written in the case that a partial failure occurred, but does not change the behaviour in the case where no updates to the indices were made. - Invoke a cleanup listener if a partial failure occurred when storing inference endpoint information in the .inference and .secrets-inference indices - Refactor ModelRegistry to use BulkRequestBuilder.add(IndexRequestBuilder) instead of the deprecated BulkRequestBuilder.add(IndexRequest) - Include cause when logging bulk failure during inference endpoint creation - Add integration tests for the new behaviour - Update docs/changelog/136577.yaml Closes elastic#123726
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Prior to this change, if only one of the .inference or .secrets-inference indices was updated when creating an inference endpoint, the endpoint creation would fail, but the successfully written doc was not removed, leading to inconsistent document counts between the two indices.
This commit removes any documents that were written in the case that a partial failure occurred, but does not change the behaviour in the case where no updates to the indices were made.
Closes #123726