Default semantic_text fields to ELSER on EIS when available #134708

mridula-s109 · 2025-09-15T09:43:22Z

🚨 Please check with search inference team on when to merge this - we're working out some rate limiting issues ahead of GA of ELSER on EIS

Summary

Implements dynamic default selection for semantic_text fields to automatically use ELSER on EIS
(.elser-2-elastic) when available, with graceful fallback to ML nodes (.elser-2-elasticsearch).

Problem

Currently, semantic_text fields are hardcoded to default to .elser-2-elasticsearch (ML nodes). This misses
opportunities for better performance and cost efficiency when EIS is available.

Solution

Dynamic Detection: Uses ModelRegistry.containsDefaultConfigId() to detect EIS availability
Smart Fallback: Automatically selects .elser-2-elastic when available, falls back to
.elser-2-elasticsearch
Zero Configuration: Works transparently without requiring user changes
Full Compatibility: On-prem users continue using ML nodes until they enable cloud-connected mode

Changes

Added getPreferredElserInferenceId() method for dynamic selection logic
Updated semantic_text field mapper to use dynamic default instead of hardcoded value
Added comprehensive tests for both EIS available/unavailable scenarios

Testing

✅ All existing tests pass (backward compatibility verified)
✅ New test covers dynamic selection logic
✅ Manual testing confirms proper fallback behavior

Impact

Cloud users: Automatically get better performance with EIS
On-prem users: No changes, continue using ML nodes seamlessly
Existing fields: Completely unaffected, no migration needed

mridula-s109 · 2025-09-15T09:59:47Z

Hey @ioanatia! 👋

Just implemented the dynamic ELSER default selection for semantic_text fields.

The approach is intentionally minimal -leverages existing ModelRegistry.containsDefaultConfigId() for
EIS detection rather than building new infrastructure. New semantic_text fields automatically default
to .elser-2-elastic when available, gracefully fall back to .elser-2-elasticsearch when not.

A few things I'd especially appreciate feedback on:

Does the overall approach make sense?
Is the error handling sufficient (graceful fallback when ModelRegistry fails)?
Any concerns about the scope or missing edge cases?

Still draft mode, but wanted to get your input before finalizing. Thanks!

...nference/src/main/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapper.java

ioanatia · 2025-09-15T15:12:44Z

relying on modelRegistry.containsDefaultConfigId(EIS_ELSER_INFERENCE_ID) seems to be right.

but I think we need more tests.
Maybe take a look at some of the tests from https://github.com/elastic/elasticsearch/tree/main/x-pack/plugin/inference/src/test/java/org/elasticsearch/xpack/inference/services/elastic and see if we can mock the presence of EIS in tests.
Then we can properly test that we pick the right inference ID - by first creating an index with a semantic_text field (and unspecified inference ID) and then requesting the mapping of this new index. We should see the default EIS endpoint.

ioanatia · 2025-09-15T15:49:26Z

In terms of tests, what we can also do is to add another mock service in https://github.com/elastic/elasticsearch/tree/main/x-pack/plugin/inference/qa/test-service-plugin/src/main/java/org/elasticsearch/xpack/inference/mock

and then use it in yaml tests.

the InferenceService allows to register default endpoints:

elasticsearch/server/src/main/java/org/elasticsearch/inference/InferenceService.java

Line 229 in 66582a3

default List<DefaultConfigId> defaultConfigIds() {

I think we can mock the default endpoints we have in EIS and use the same name in the mock inference service.
Let me know if you need any help!

...rc/main/java/org/elasticsearch/xpack/inference/services/elastic/ElasticInferenceService.java

mridula-s109 · 2025-10-14T19:49:23Z

FYI folks we're working on some rate limiting issues with ELSER and we may want to defer this.

Please check with @maxjakob or myself before merging.

Thanks @Mikep86! All comments addressed. Have verified about the default assertion in the SemanticTextEISDefaultIT, there is no flakiness or underlying issue at the moment. Since we're holding off on merging due to the ELSER rate limiting work, no rush on further review. Take your time! 😊

...nce/src/test/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapperTests.java

...e-tests/src/javaRestTest/java/org/elasticsearch/xpack/inference/InferenceSemanticTextIT.java

...-tests/src/javaRestTest/java/org/elasticsearch/xpack/inference/SemanticTextEISDefaultIT.java

mridula-s109 · 2025-10-17T11:31:25Z

@ioanatia @Mikep86 quick update: I’ve addressed the review feedback and pushed the latest changes.
Just a heads-up we’ve scheduled a short sync with @seanhandley and @maxjakob on Wednesday, October 22 (1:00–1:15 PM) to proceed with merging this PR.
Please let me know if there’s anything else you’d like me to adjust before then. 🙌

Mikep86

LGTM to merge once we get the 👍 from @seanhandley and @maxjakob

...-tests/src/javaRestTest/java/org/elasticsearch/xpack/inference/SemanticTextEISDefaultIT.java

liranabn · 2025-10-20T10:42:41Z

@mridula-s109 , do we have a separate PR to update the docs?
specifically here

If you don’t specify an inference endpoint, the inference_id field defaults to .elser-2-elasticsearch, a preconfigured endpoint for the elasticsearch service.

seanhandley · 2025-10-21T14:18:57Z

Great work @mridula-s109 !

We've decided to wait anther week while we investigate a bug report on our side. Once it's mitigated we can hopefully merge this. I've pushed our calendar reminder back to next week to accommodate this.

mridula-s109 · 2025-10-21T14:43:32Z

Great work @mridula-s109 !

We've decided to wait anther week while we investigate a bug report on our side. Once it's mitigated we can hopefully merge this. I've pushed our calendar reminder back to next week to accommodate this.

Thanks for the update @seanhandley! 👍

No problem at all, waiting another week makes sense to ensure everything is stable. I'll keep an eye out for next week's calendar reminder.

In the meantime, @liranabn raised a good point about doc updates. I'll use this week to add the documentation changes directly to this PR (updating the section that mentions .elser-2-elasticsearch as the default). Will push those updates in the next day or two.

mridula-s109 · 2025-10-21T15:04:50Z

@mridula-s109 , do we have a separate PR to update the docs? specifically here

If you don’t specify an inference endpoint, the inference_id field defaults to .elser-2-elasticsearch, a preconfigured endpoint for the elasticsearch service.

Are there any other doc pages or files that reference the .elser-2-elasticsearch default that I should also update?@Mikep86 @ioanatia Want to make sure I catch all the places that need changes. Thanks!

leemthompo · 2025-10-21T15:13:09Z

@mridula-s109

Istvan already has a PR lined up for the elasticsearch repo, which you can check out: Adds ELSER on EIS examples to semantic_text #135359
there are a couple of pages in the docs-content repo:

One of these is a semantic_text tutorial being handled in a PR already ([SEARCH] Adds EIS scenario to semantic search tutorial docs-content#3146).
The other is https://github.com/elastic/docs-content/blob/main/explore-analyze/elastic-inference/inference-api.md?plain=1#L37 which you might be able to handle. :)

I also see one instance in the spec repo: https://github.com/elastic/elasticsearch-specification/blob/main/specification/_types/mapping/core.ts#L246

mridula-s109 · 2025-10-21T15:35:18Z

@mridula-s109

Istvan already has a PR lined up for the elasticsearch repo, which you can check out: Adds ELSER on EIS examples to semantic_text #135359

there are a couple of pages in the docs-content repo:

One of these is a semantic_text tutorial being handled in a PR already ([SEARCH] Adds EIS scenario to semantic search tutorial docs-content#3146).

The other is https://github.com/elastic/docs-content/blob/main/explore-analyze/elastic-inference/inference-api.md?plain=1#L37 which you might be able to handle. :)

I also see one instance in the spec repo: https://github.com/elastic/elasticsearch-specification/blob/main/specification/_types/mapping/core.ts#L246

Thanks @leemthompo for the inputs, i will look into it!

...nference/src/main/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapper.java

seanhandley · 2025-10-29T10:27:15Z

@mridula-s109 We're clear to merge this today 👍 There are other pieces of work we want to do but none of them are blocking this from being merged.

https://github.com/elastic/search-team/issues/11650

cc @sphilipse @liranabn @maxjakob

mridula-s109 · 2025-10-29T10:38:39Z

@mridula-s109 We're clear to merge this today 👍 There are other pieces of work we want to do but none of them are blocking this from being merged.

elastic/search-team#11650

cc @sphilipse @liranabn @maxjakob

Thanks for letting me know @seanhandley, As the CI has passed as well post updating from main today, i will go ahead with the merge now.

mridula-s109 · 2025-10-29T10:47:11Z

Opening up a followup PR to modify the SemanticFieldMappertests to point to the new constant in SemanticTextFieldMapper.java.

…#134708) * Defaulting EIS on ELSER * Extended testing * [CI] Auto commit changes from spotless * Edited to include the default * Cleaned up the mapper implementation * COmpile issue * [CI] Auto commit changes from spotless * Added tests * [CI] Auto commit changes from spotless * Refactored the variable names * Cleanup done * Removed unnecessary files * Unit tests and mock is working * [CI] Auto commit changes from spotless * Fix test * yaml addition failure * [CI] Auto commit changes from spotless * Resolved the import issue or duplication of variables in mock * Resolved PR comments * Restored error * Update docs/changelog/134708.yaml * Integration test * [CI] Auto commit changes from spotless * Resolved all PR comments * [CI] Auto commit changes from spotless * Cleaned up the redudant reference of TestInferencePlugin * Included both before and before test * [CI] Auto commit changes from spotless * Made changes to accomodate the old constant version string --------- Co-authored-by: elasticsearchmachine <[email protected]>

Defaulting EIS on ELSER

328d079

mridula-s109 requested a review from ioanatia September 15, 2025 09:43

mridula-s109 self-assigned this Sep 15, 2025

mridula-s109 added >enhancement v9.2.0 labels Sep 15, 2025

mridula-s109 and others added 2 commits September 15, 2025 10:52

Extended testing

0b89373

[CI] Auto commit changes from spotless

1cc9fd4

ioanatia reviewed Sep 15, 2025

View reviewed changes

...nference/src/main/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapper.java Outdated Show resolved Hide resolved

ioanatia reviewed Sep 15, 2025

View reviewed changes

...nference/src/main/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapper.java Outdated Show resolved Hide resolved

mridula-s109 added 4 commits September 17, 2025 16:18

Merge branch 'main' into default_elser_on_eis_semantic

0829058

Edited to include the default

7dcd31f

Cleaned up the mapper implementation

f19972e

COmpile issue

3d58a50

mridula-s109 commented Sep 17, 2025

View reviewed changes

...rc/main/java/org/elasticsearch/xpack/inference/services/elastic/ElasticInferenceService.java Show resolved Hide resolved

elasticsearchmachine and others added 5 commits September 17, 2025 15:47

[CI] Auto commit changes from spotless

394bf99

Merge branch 'main' into default_elser_on_eis_semantic

213aa24

Added tests

c92d2b3

[CI] Auto commit changes from spotless

1301a43

Merge branch 'main' into default_elser_on_eis_semantic

6ecdb6c

elasticsearchmachine added v9.3.0 and removed v9.2.0 labels Oct 2, 2025

mridula-s109 and others added 6 commits October 3, 2025 00:24

Refactored the variable names

9f662ce

Cleanup done

904a242

Removed unnecessary files

6355bed

Unit tests and mock is working

b4ed763

[CI] Auto commit changes from spotless

4f0d7c0

Merge branch 'main' into default_elser_on_eis_semantic

4680417

[CI] Auto commit changes from spotless

dd765ae

mridula-s109 requested review from Mikep86 and kderusso October 14, 2025 19:56

Merge branch 'main' into default_elser_on_eis_semantic

5d932ef

Mikep86 reviewed Oct 15, 2025

View reviewed changes

...nce/src/test/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapperTests.java Outdated Show resolved Hide resolved

...e-tests/src/javaRestTest/java/org/elasticsearch/xpack/inference/InferenceSemanticTextIT.java Outdated Show resolved Hide resolved

mridula-s109 added 2 commits October 17, 2025 12:10

Cleaned up the redudant reference of TestInferencePlugin

536d326

Included both before and before test

cd98727

mridula-s109 commented Oct 17, 2025

View reviewed changes

...-tests/src/javaRestTest/java/org/elasticsearch/xpack/inference/SemanticTextEISDefaultIT.java Show resolved Hide resolved

elasticsearchmachine and others added 2 commits October 17, 2025 11:18

[CI] Auto commit changes from spotless

cf797e4

Merge branch 'main' into default_elser_on_eis_semantic

9e50bdc

Mikep86 approved these changes Oct 17, 2025

View reviewed changes

...-tests/src/javaRestTest/java/org/elasticsearch/xpack/inference/SemanticTextEISDefaultIT.java Show resolved Hide resolved

ioanatia approved these changes Oct 20, 2025

View reviewed changes

mridula-s109 added 3 commits October 27, 2025 18:59

Merge branch 'main' into default_elser_on_eis_semantic

d5c4841

Made changes to accomodate the old constant version string

f059474

Merge branch 'main' into default_elser_on_eis_semantic

bc7a73a

mridula-s109 commented Oct 29, 2025

View reviewed changes

...nference/src/main/java/org/elasticsearch/xpack/inference/mapper/SemanticTextFieldMapper.java Show resolved Hide resolved

mridula-s109 merged commit 6b2c55a into elastic:main Oct 29, 2025
34 checks passed

mridula-s109 mentioned this pull request Oct 29, 2025

Remove deprecated constant for default elser inference id #137329

Merged

Default semantic_text fields to ELSER on EIS when available #134708

Default semantic_text fields to ELSER on EIS when available #134708

Uh oh!

Conversation

mridula-s109 commented Sep 15, 2025 • edited by seanhandley Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Problem

Solution

Changes

Testing

Impact

Uh oh!

mridula-s109 commented Sep 15, 2025

Uh oh!

Uh oh!

Uh oh!

ioanatia commented Sep 15, 2025

Uh oh!

ioanatia commented Sep 15, 2025

Uh oh!

Uh oh!

mridula-s109 commented Oct 14, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

mridula-s109 commented Oct 17, 2025

Uh oh!

Mikep86 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

liranabn commented Oct 20, 2025

Uh oh!

seanhandley commented Oct 21, 2025

Uh oh!

mridula-s109 commented Oct 21, 2025

Uh oh!

mridula-s109 commented Oct 21, 2025

Uh oh!

leemthompo commented Oct 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mridula-s109 commented Oct 21, 2025

Uh oh!

Uh oh!

seanhandley commented Oct 29, 2025

Uh oh!

mridula-s109 commented Oct 29, 2025

Uh oh!

Uh oh!

mridula-s109 commented Oct 29, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

mridula-s109 commented Sep 15, 2025 •

edited by seanhandley

Loading

leemthompo commented Oct 21, 2025 •

edited

Loading