excluding circuit breaker for Agent by dhrubo-os · Pull Request #3814 · opensearch-project/ml-commons

dhrubo-os · 2025-04-30T15:51:13Z

Description

[excluding circuit breaker for Agent]

Related Issues

Resolves #[Issue number to be closed when this PR is merged]

Check List

New functionality includes testing.
New functionality has been documented.
API changes companion pull request created.
Commits are signed per the DCO using --signoff.
Public documentation issue/PR created.

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com>

b4sjoo · 2025-04-30T18:01:26Z

plugin/src/main/java/org/opensearch/ml/task/MLTaskRunner.java

                log.debug("Execute ML request {} locally on node {}", request.getRequestID(), nodeId);
-                checkOpenCircuitBreaker(mlCircuitBreakerService, mlStats);
-                executeTask(request, listener);
+                checkCBAndExecute(functionName, request, listener);


Nit: Personally I don't like abbreviation. Try to use CircuitBreaker instead of CB?

This is already an existing method. I'm leaving it as it now.

are you using the latest main branch? I think this change is redundant in the main as checkCBAndExecute is already there?

public void run(FunctionName functionName, Request request, TransportService transportService, ActionListener<Response> listener) { if (!request.isDispatchTask()) { log.debug("Run ML request {} locally", request.getRequestID()); checkCBAndExecute(functionName, request, listener); return; } dispatchTask(functionName, request, transportService, listener); }

This is where we applied.

I'm using latest main branch.

Zhangxunmt · 2025-04-30T18:21:40Z

plugin/src/main/java/org/opensearch/ml/task/MLTaskRunner.java

                log.debug("Execute ML request {} locally on node {}", request.getRequestID(), nodeId);
-                checkOpenCircuitBreaker(mlCircuitBreakerService, mlStats);
-                executeTask(request, listener);
+                checkCBAndExecute(functionName, request, listener);


are you using the latest main branch? I think this change is redundant in the main as checkCBAndExecute is already there?

Zhangxunmt · 2025-04-30T18:23:07Z

plugin/src/main/java/org/opensearch/ml/model/MLModelManager.java

-        if (Objects.nonNull(mlTask) && mlTask.getFunctionName() != FunctionName.REMOTE) {
+
+        // for agent and remote model prediction we don't need to check circuit breaker
+        if (Objects.nonNull(mlTask) && mlTask.getFunctionName() != FunctionName.REMOTE && mlTask.getFunctionName() != FunctionName.AGENT) {


For agent using pre-trained model, does this skip the CB?

That a good question. It will skip agent level execution. But under the agent execution when a model prediction is going to happen, then function name is not remote or agent, so that time it will check CB.

Sure we can merge for now but need to remember that this would cause problems when agent is not using a remote model.

For old versions, can user update setting to disable circuit breaker checking ?

From code, looks like we only check memory and disk.

So if somebody sets up, jvm_heap_memory_threshold to 100 & disk_free_space_threshold to -1 then circuit breaker will be disabled.

Does that answer your question @ylwu-amzn ?

Sounds good, @dhrubo-os , can you check if we have this in our document?

which document are you referring to? I shared the memory and disk link from docs.opensearch.org.

codecov · 2025-04-30T19:33:57Z

Codecov Report

Attention: Patch coverage is 66.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 78.00%. Comparing base (ed4f09f) to head (1701c23).
Report is 8 commits behind head on main.

Files with missing lines	Patch %	Lines
...n/java/org/opensearch/ml/model/MLModelManager.java	0.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #3814      +/-   ##
============================================
- Coverage     78.20%   78.00%   -0.21%     
- Complexity     7162     7317     +155     
============================================
  Files           631      655      +24     
  Lines         32248    32997     +749     
  Branches       3666     3708      +42     
============================================
+ Hits          25219    25738     +519     
- Misses         5465     5673     +208     
- Partials       1564     1586      +22

Flag	Coverage Δ
ml-commons	`78.00% <66.66%> (-0.21%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> (cherry picked from commit f01de7f)

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> (cherry picked from commit f01de7f) Co-authored-by: Dhrubo Saha <dhrubo@amazon.com>

* [BUG] Agent Framework: Handle model response when toolUse is not accompanied by text (#3755) * fix: handle model response when toolUse is not accompanied by text Signed-off-by: Pavan Yekbote <pybot@amazon.com> * feat: add test case for parseLLMOutput Signed-off-by: Pavan Yekbote <pybot@amazon.com> --------- Signed-off-by: Pavan Yekbote <pybot@amazon.com> * [BUG] Allow user to control react agent max_interations value to prevent empty response (#3756) * fix: expose max_iteration for react Signed-off-by: Pavan Yekbote <pybot@amazon.com> * fix: defaults for agent execution and differentiate between step and step result Signed-off-by: Pavan Yekbote <pybot@amazon.com> * fix: return react agent id in agent response to expose more details Signed-off-by: Pavan Yekbote <pybot@amazon.com> * spotless Signed-off-by: Pavan Yekbote <pybot@amazon.com> * fix: remove test prompt from react system prompt Signed-off-by: Pavan Yekbote <pybot@amazon.com> * refactor: rename parameters exposed to user to executor Signed-off-by: Pavan Yekbote <pybot@amazon.com> * fix: give user complete control over planner system prompt Signed-off-by: Pavan Yekbote <pybot@amazon.com> --------- Signed-off-by: Pavan Yekbote <pybot@amazon.com> * Clean up JSM from MCP (#3773) Signed-off-by: rithin-pullela-aws <rithinp@amazon.com> * [Bug] ListTools call does not return tool attributes (#3785) * initial commit for MCP server in OpenSearch (#3781) * initial commit for MCP server in OpenSearch Signed-off-by: zane-neo <zaniu@amazon.com> * Make change to support register or remove tools across cluster Signed-off-by: zane-neo <zaniu@amazon.com> * format code Signed-off-by: zane-neo <zaniu@amazon.com> * fix UT failure caused by code change Signed-off-by: zane-neo <zaniu@amazon.com> * format code Signed-off-by: zane-neo <zaniu@amazon.com> * format code Signed-off-by: zane-neo <zaniu@amazon.com> * add license header Signed-off-by: zane-neo <zaniu@amazon.com> * fix notifications initialized not respond issue Signed-off-by: zane-neo <zaniu@amazon.com> * fix minor issues and add UTs Signed-off-by: zane-neo <zaniu@amazon.com> * Add more UTs Signed-off-by: zane-neo <zaniu@amazon.com> --------- Signed-off-by: zane-neo <zaniu@amazon.com> * Remove beta1 qualifier (#3794) (#3795) (cherry picked from commit 3f503f1) Signed-off-by: Peter Zhu <zhujiaxi@amazon.com> Co-authored-by: Peter Zhu <zhujiaxi@amazon.com> * [AUTO] Increment version to 3.1.0-SNAPSHOT (#3789) * Increment version to 3.1.0-SNAPSHOT Signed-off-by: opensearch-ci-bot <opensearch-infra@amazon.com> * Update build.gradle Signed-off-by: Peter Zhu <zhujiaxi@amazon.com> --------- Signed-off-by: opensearch-ci-bot <opensearch-infra@amazon.com> Signed-off-by: Peter Zhu <zhujiaxi@amazon.com> Co-authored-by: opensearch-ci-bot <opensearch-infra@amazon.com> Co-authored-by: Peter Zhu <zhujiaxi@amazon.com> * add release note for 3.0 (#3792) Signed-off-by: Mingshi Liu <mingshl@amazon.com> * support MCP session management (#3803) * support MCP session management Signed-off-by: zane-neo <zaniu@amazon.com> * Addressing comments Signed-off-by: zane-neo <zaniu@amazon.com> * add feature flag for mcp server and renaming mcp connector feature flag Signed-off-by: zane-neo <zaniu@amazon.com> * Address critical comments in #3781 Signed-off-by: zane-neo <zaniu@amazon.com> --------- Signed-off-by: zane-neo <zaniu@amazon.com> * upgrade http client to version align with core (#3809) * upgrade http client to versoin align with core Signed-off-by: zane-neo <zaniu@amazon.com> * upgrade httpclient-h2 to correct versiono Signed-off-by: zane-neo <zaniu@amazon.com> * use placeholder approach Signed-off-by: zane-neo <zaniu@amazon.com> --------- Signed-off-by: zane-neo <zaniu@amazon.com> * support customized message endpoint and addressing comments (#3810) * support customized message endpoint and addressing comments Signed-off-by: zane-neo <zaniu@amazon.com> * fix UT failures Signed-off-by: zane-neo <zaniu@amazon.com> * add files to jacoco exception Signed-off-by: zane-neo <zaniu@amazon.com> * fix tool name issue and optimize register tool api Signed-off-by: zane-neo <zaniu@amazon.com> * fix schema not parsed correctly issue and NPE when parameters is null Signed-off-by: zane-neo <zaniu@amazon.com> * fix failure UT Signed-off-by: zane-neo <zaniu@amazon.com> --------- Signed-off-by: zane-neo <zaniu@amazon.com> * excluding circuit breaker for Agent (#3814) Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> * change release note (#3811) * change release note Signed-off-by: zane-neo <zaniu@amazon.com> * Update opensearch-ml-common.release-notes-3.0.0.0.md * Update opensearch-ml-common.release-notes-3.0.0.0.md * Update opensearch-ml-common.release-notes-3.0.0.0.md --------- Signed-off-by: zane-neo <zaniu@amazon.com> Co-authored-by: Peter Zhu <zhujiaxi@amazon.com> * Downgrade MCP version to 0.9 (#3821) Signed-off-by: rithin-pullela-aws <rithinp@amazon.com> * remove libs folder (#3824) Signed-off-by: Yaliang Wu <ylwu@amazon.com> * add more logging to deploy/undeploy flows for better debugging (#3825) * add more logging to deploy/undeploy flows for better debugging Signed-off-by: Bhavana Goud Ramaram <rbhavna@amazon.com> * Fix python client not able to connect to MCP server issue (#3822) Signed-off-by: zane-neo <zaniu@amazon.com> Co-authored-by: Dhrubo Saha <dhrubo@amazon.com> * exclude trusted connector check for hidden model (#3838) Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> * adding tenantId to the connector executor when this is inline connector (#3837) * adding tenantId to the connector executor when this is inline connector Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> * added more unit tests Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> --------- Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> --------- Signed-off-by: Pavan Yekbote <pybot@amazon.com> Signed-off-by: rithin-pullela-aws <rithinp@amazon.com> Signed-off-by: zane-neo <zaniu@amazon.com> Signed-off-by: Peter Zhu <zhujiaxi@amazon.com> Signed-off-by: opensearch-ci-bot <opensearch-infra@amazon.com> Signed-off-by: Mingshi Liu <mingshl@amazon.com> Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> Signed-off-by: Yaliang Wu <ylwu@amazon.com> Signed-off-by: Bhavana Goud Ramaram <rbhavna@amazon.com> Co-authored-by: Pavan Yekbote <pybot@amazon.com> Co-authored-by: Rithin Pullela <rithinp@amazon.com> Co-authored-by: zane-neo <zaniu@amazon.com> Co-authored-by: opensearch-trigger-bot[bot] <98922864+opensearch-trigger-bot[bot]@users.noreply.github.com> Co-authored-by: Peter Zhu <zhujiaxi@amazon.com> Co-authored-by: opensearch-ci-bot <opensearch-infra@amazon.com> Co-authored-by: Mingshi Liu <mingshl@amazon.com> Co-authored-by: Yaliang Wu <ylwu@amazon.com> Co-authored-by: Bhavana Goud Ramaram <rbhavna@amazon.com>

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> Signed-off-by: Abdul Muneer Kolarkunnu <muneer.kolarkunnu@netapp.com>

dhrubo-os requested review from HenryL27, Zhangxunmt, austintlee, b4sjoo, jngz-es, mingshl, model-collapse, rbhavna, xinyual, ylwu-amzn and zane-neo as code owners April 30, 2025 15:51

dhrubo-os force-pushed the circuit_breaker branch from 747b4d6 to a4eb8fd Compare April 30, 2025 15:55

dhrubo-os had a problem deploying to ml-commons-cicd-env April 30, 2025 15:56 — with GitHub Actions Error

dhrubo-os had a problem deploying to ml-commons-cicd-env April 30, 2025 15:56 — with GitHub Actions Failure

peterzhuamazon added the backport 3.0 label Apr 30, 2025

excluding circuit breaker for Agent

1701c23

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com>

dhrubo-os force-pushed the circuit_breaker branch from a4eb8fd to 1701c23 Compare April 30, 2025 17:54

dhrubo-os had a problem deploying to ml-commons-cicd-env April 30, 2025 17:55 — with GitHub Actions Error

dhrubo-os had a problem deploying to ml-commons-cicd-env April 30, 2025 17:55 — with GitHub Actions Failure

dhrubo-os had a problem deploying to ml-commons-cicd-env April 30, 2025 17:55 — with GitHub Actions Error

dhrubo-os had a problem deploying to ml-commons-cicd-env April 30, 2025 17:55 — with GitHub Actions Failure

peterzhuamazon mentioned this pull request Apr 30, 2025

[RELEASE] Release version 3.0.0 opensearch-project/opensearch-build#3747

Closed

66 tasks

b4sjoo reviewed Apr 30, 2025

View reviewed changes

rbhavna approved these changes Apr 30, 2025

View reviewed changes

b4sjoo approved these changes Apr 30, 2025

View reviewed changes

jngz-es approved these changes Apr 30, 2025

View reviewed changes

Zhangxunmt reviewed Apr 30, 2025

View reviewed changes

dhrubo-os temporarily deployed to ml-commons-cicd-env April 30, 2025 18:31 — with GitHub Actions Inactive

dhrubo-os temporarily deployed to ml-commons-cicd-env April 30, 2025 19:34 — with GitHub Actions Inactive

dhrubo-os merged commit f01de7f into opensearch-project:main Apr 30, 2025
12 of 14 checks passed

opensearch-trigger-bot bot pushed a commit that referenced this pull request Apr 30, 2025

excluding circuit breaker for Agent (#3814)

b47a9d5

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> (cherry picked from commit f01de7f)

opensearch-trigger-bot bot mentioned this pull request Apr 30, 2025

[Backport 3.0] excluding circuit breaker for Agent #3818

Merged

peterzhuamazon pushed a commit that referenced this pull request Apr 30, 2025

excluding circuit breaker for Agent (#3814) (#3818)

7fa69e3

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> (cherry picked from commit f01de7f) Co-authored-by: Dhrubo Saha <dhrubo@amazon.com>

akolarkunnu pushed a commit to akolarkunnu/ml-commons that referenced this pull request Jun 6, 2025

excluding circuit breaker for Agent (opensearch-project#3814)

240d6b0

Signed-off-by: Dhrubo Saha <dhrubo@amazon.com> Signed-off-by: Abdul Muneer Kolarkunnu <muneer.kolarkunnu@netapp.com>

Conversation

dhrubo-os commented Apr 30, 2025

Description

Related Issues

Check List

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codecov bot commented Apr 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

codecov bot commented Apr 30, 2025 •

edited

Loading