test: reduce sidecar testing flakes - #4186
Conversation
|
There was a problem hiding this comment.
Is there a specific reason you dropped the debuggability of this test? Otherwise probably fine.
There was a problem hiding this comment.
I guess you consider hexdumps more debuggable than messages? xD
I can revert that aspect, sure.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
Description
This fixes two different sidecar flakes.
inWebServer()retries the entire server lifecycle when no traces are received, buttestComposerInteroperabilityWhenNoInitHookexpects no traces. It then kills a healthy PHP server and immediately restarts it on the same port. With telemetry enabled, the extension can start a sidecar that briefly outlives the forcibly stopped PHP process and may retain the server socket. The unnecessary retry then fails with Server never came up. Fixed by settingretries=0for this test. There are no other tests I could find that also assert that no traces are made.agent_sampling_sidecar.phptcaused by matching stale shared-memory data from earlier or parallel test processes. Each invocation now uses unique sampling configuration markers, ensuringit waits for its own sidecar updates. Failures also report concise, actionable messages instead of dumping large shared-memory buffers.
Reviewer checklist