Document automatic vector exclusion from _source (knn_default_excludes) - #12979
Document automatic vector exclusion from _source (knn_default_excludes)#12979shatejas wants to merge 1 commit into
Conversation
|
Thank you for submitting your PR. The PR states are In progress (or Draft) -> Tech review -> Doc review -> Merged. Before you submit your PR for doc review, make sure the content is technically accurate. If you need help finding a tech reviewer, tag a maintainer. When you're ready for doc review, tag the assignee of this PR. The doc reviewer may push edits to the PR directly or leave comments and editorial suggestions for you to address (let us know in a comment if you have a preference). |
|
@kolchfa-aws Can we put a backport 3.8 label for this please, Thanks! |
Add documentation for the knn_default_excludes system-generated search request processor, introduced in OpenSearch 3.8, which automatically excludes knn_vector fields from _source in search responses. - Add an "Automatically exclude vectors from search results" section to the vector search performance tuning page, covering how to enable the processor, default behavior, how to override it, and a note that scoring is unaffected because exclusion is applied on the fetch path. - Register knn_default_excludes in the system-generated search request processors table and add a related-documentation link. Signed-off-by: Tejas Shah <shatejas@amazon.com>
b114729 to
ca54e25
Compare
|
|
||
| Instead of manually specifying vector fields in `_source.excludes` on every request, you can let OpenSearch exclude them automatically. When enabled, the `knn_default_excludes` [system-generated search request processor]({{site.url}}{{site.baseurl}}/search-plugins/search-pipelines/system-generated-search-processors/) inspects the index mappings for each search request, finds all `knn_vector` fields (including those nested in object or nested fields), and adds them to `_source.excludes` before the request runs. This reduces the search response payload without requiring any per-request configuration. | ||
|
|
||
| The processor is disabled by default. To enable it, add its factory, `knn_default_excludes_factory`, to the `cluster.search.enabled_system_generated_factories` cluster setting: |
There was a problem hiding this comment.
Why is it disabled by default? I thought we want to enable by default
There was a problem hiding this comment.
There isn't a way to enable it by default, users can do that in opensearch.yml they create
Add documentation for the knn_default_excludes system-generated search request processor, introduced in OpenSearch 3.8, which automatically excludes knn_vector fields from _source in search responses.
Description
Describe what this change achieves.
Issues Resolved
Closes #[Replace this text, including the brackets, with the issue number. Leave "Closes #" so the issue is closed properly.]
Version
List the OpenSearch version to which this PR applies, e.g. 2.14, 2.12--2.14, or all.
Frontend features
If you're submitting documentation for an OpenSearch Dashboards feature, add a video that shows how a user will interact with the UI step by step. A voiceover is optional.
Checklist
For more information on following Developer Certificate of Origin and signing off your commits, please check here.