deytenit/apex.ermnvldmr.com
1
0 Fork 0
Code Issues Pull Requests Actions Releases Wiki Activity

fix(readiness): allow measured native probe execution time

Browse Source
This commit is contained in:
deytenit committed 2026-10-11 21:30:14 +03:00
1 parent 31938681dc
commit e4308a2dd4
5 files changed
+18 -8
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

assets/composition-readiness/README.md
+1 -1
View File
@@ -21,7 +21,7 @@ Run `python3 adapter.py --config /private/adapter.json` inside the application f
Only GET at the exact canonical path, with one matching `X-APEX-Ready-Token` header, can execute the fixed command. Missing/wrong/duplicate credentials return403; other paths404; supported non-GET methods405. Ready returns204; probe failure, timeout or concurrency exhaustion returns503. These responses have empty bodies and `Cache-Control: no-store`. The adapter does not log requests, tokens or command output, cache success, or accept a client-supplied command. Each connection closes after its response.
Probe timeout is bounded to1.5seconds so an origin can answer within the Worker's two-second deadline. A timeout kills the probe's process group. Probe slots and request threads are bounded; slow socket operations time out after two seconds. Operators must account for proxy overhead and actual probe execution time. Probe programs must not detach into a new session or launch background jobs.
Probe timeout defaults to1.5seconds and can be configured up to3seconds. With a3-second probe, allow4seconds at the readiness proxy and5seconds at the Worker. A timeout kills the probe's process group. Probe slots and request threads are bounded; slow socket operations time out after two seconds. Operators must account for proxy overhead and actual probe execution time. Probe programs must not detach into a new session or launch background jobs.
Optional `single_flight: true` lets overlapping authenticated requests share one running fixed-command probe. It accepts only a JSON boolean and defaults to false. Waiters use that probe's remaining monotonic deadline; they never queue another probe for the same request. The existing connection limit (`slots * 2 + 4`) also bounds waiting requests. Timeout or shutdown returns unavailable and terminates the probe group. Completed results are not cached: each subsequent request starts a fresh probe, with no success TTL or failure cache. Coalescing reduces concurrent work but does not make a slow individual probe meet its deadline.