chore(e2e): remove tests/e2e/docker-compose.yml - #33837
Conversation
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
🤖 Devin AI EngineerI'll be helping with this pull request! Here's what you should know: ✅ I will automatically:
Note: I can only respond to comments from users who have write access to this repository. ⚙️ Control Options:
|
|
|
Greptile SummaryThis PR deletes
Confidence Score: 3/5Merging this breaks the local e2e development workflow for all contributors until a follow-up lands; no production code is affected. The deletion removes the only local-dev entrypoint for the entire e2e suite. CONTRIBUTING.md, CLAUDE.md, e2e_config.py, and multiple test files all reference this compose file by name. Any contributor following the documented setup steps after this merges will immediately hit a missing file. The PR itself acknowledges the dangling references and defers cleanup to a follow-up, but until that follow-up lands the e2e directory is in an inconsistent state — the docs say 'run docker compose up -d' and the file simply isn't there. tests/e2e/docker-compose.yml is the only changed file; the concern is the downstream impact on CONTRIBUTING.md, CLAUDE.md, e2e_config.py, and several test conftest files that all reference it.
|
| Filename | Overview |
|---|---|
| tests/e2e/docker-compose.yml | Deletes the e2e harness's compose stack (Postgres, Redis, Jaeger, FastMCP upstream, LiteLLM proxy), leaving docker compose up -d from tests/e2e/ broken and all references in CONTRIBUTING.md, CLAUDE.md, e2e_config.py, and multiple test files dangling. |
Comments Outside Diff (1)
-
tests/e2e/docker-compose.ymlDangling references break local e2e setup
Removing this file leaves the contributor workflow in a broken state.
CONTRIBUTING.mdinstructs contributors to rundocker compose up -dfromtests/e2e/as step 2 of local setup, and step 2 of the pre-commit checklist says "Add the models your test needs to the inline config indocker-compose.yml".CLAUDE.mdhas a hard rule: "use the docker-compose to your advantage and spin up a local proxy, make sure all tests pass."e2e_config.pyhas an inline comment tyingOTEL_QUERY_URLto the compose stack'sjaegerservice port mapping. Several test conftest files and budget-reset tests also referencedocker compose. Without a replacement entrypoint landing in the same commit, any contributor following the documented setup steps will hit a missing file immediately.Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!
Reviews (1): Last reviewed commit: "chore(e2e): remove tests/e2e/docker-comp..." | Re-trigger Greptile
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
…r_compose Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Adapt the Azure OpenAI chat coverage onto the refactored e2e harness: drop tests/e2e/docker-compose.yml (removed upstream in #33837) and register the Azure deployments via /model/new inside the test with teardown, keep the upstream stream_events transport and add a stream_done terminator flag, port the ChatStreamChunk models, and extend LiteLLMParamsBody/ModelInfoBody with the max_tokens, drop_params, and base_model fields the Azure cases need Also drop the strict xfail from test_azure_config_token_cap_with_client_max_completion_tokens now that GH #31614 is fixed: AzureOpenAIConfig.map_openai_params skips max_tokens when the client also sends max_completion_tokens (PR #34214), so the row asserts the completion succeeds and guards against regression
Relevant issues
Linear ticket
Pre-Submission checklist
Please complete all items before asking a LiteLLM maintainer to review your PR
@greptileaito re-request a review after pushing changes)Type
🧹 Refactoring
Changes
Deletes
tests/e2e/docker-compose.ymland repoints the contributor docs at running the litellm proxy locally instead ofdocker compose upThe deleted compose was the e2e harness's stack, not a duplicate of the root
docker-compose.yml(the root one is a general local-dev deploy: builds from source, persistent Postgres, Prometheus, no inline proxy config or models). With it gone,tests/e2e/CONTRIBUTING.mdandtests/e2e/CLAUDE.mdnow tell contributors to stand up a Postgres and Redis, point a.envat them, and runlitellm --config <your-e2e-config>.yml --port 4000, then run the suite againsthttp://localhost:4000. The docs also call out the services individual suites still need (a jaeger for the OTEL logging tests, the upstream MCP server for the mcp suite)Scope note: this does not ship a replacement e2e proxy config, so contributors supply their own config wiring the example models the suites read back (
gpt-5.5,claude-haiku-4-5,gemini-2.5-flash,openai-text-embedding-3-small,complexity-smart-router). Inline code comments in a few tests still mention the olddocker-compose.yml(e2e_config.py,router/conftest.py,router/test_complexity_router_e2e.py,management/test_key_models_dropdown_e2e.py,claude_code/conftest.py, the budget-window tests); left untouched here to keep the change to the two contributor docsFinal Attestation
Link to Devin session: https://app.devin.ai/sessions/60e23180816b42af9de38382c30e2ccc
Requested by: @yassin-berriai