Add durable MLflow agent server - #468
Closed
shivam5 wants to merge 3 commits into
Closed
Conversation
This was referenced Aug 19, 2026
Contributor
Author
|
This is a POC PR - moved the PR to #467 |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What did you change, and why?
Change: Add
DatabricksDurableAgentServer, an MLflowAgentServersubclass backed byDatabricksDurableRuntime. It preserves the standard@invoke()handler, AgentServer validation/tracing/header propagation/chat proxy, and adds blocking/background execution, Lakebase recovery, and result retrieval. Adaptexamples/openai-sdk-agentto use this abstraction and document the DX and inheritance tradeoff.Why: This is a sibling design alternative to #466 and #467 for comparing a transport-neutral runtime, a standalone composed server, and an AgentServer-derived server. The application no longer owns FastAPI or status-response glue; the library accepts tighter coupling to AgentServer's protected route and invoke hooks.
How do you know it works?
Testing: Added tests for inheritance, AgentServer validation,
@invoke()dispatch, durable attempt context, blocking/background/retrieval/cache/error behavior, lifecycle, health, and/agent/info. The focused server and durable-runtime suites pass (30 tests) against both MLflow 3.10.1 and 3.15.1. Repository formatting and lint pass;ty checkcompletes with three pre-existing unused-ignore warnings. The included live stop/start observations are explicitly identified as results from the shared runtime baseline; this transport variant was not separately deployed.