He-Pin opened a new pull request, #3090:
URL: https://github.com/apache/pekko/pull/3090

   ### Motivation
   
   The "must be able to send messages with actorSelection concurrently 
preserving order" test flakes on CI: 4 sender actors each drive 1000 round 
trips via `ActorSelection` (4004 messages total), wrapped in 
`ActorSelectionMessage` with remote-side path traversal overhead. On CI with 
`-Dpekko.test.timefactor=2`, `within(30.seconds)` dilates to 60s which is 
insufficient under load — 3 of 4 senders complete but the 4th times out waiting 
for the final `success2` message.
   
   Investigation confirmed no recent code change caused this:
   - Artery does not use stage actors → `LazyDispatch` (ba4e950edb) has zero 
impact
   - Materializer wiring optimization (03ebaf59bb) affects materialization 
only, not runtime message flow
   - `GraphInterpreter` `pendingFinalization` (a5058437af) is a hot-path 
optimization, no behavioral change
   
   ### Modification
   
   Bump `within(30.seconds)` → `within(60.seconds)` at line 219 for the 
ActorSelection test only. The ActorRef variant (line 179) is unchanged as it 
passes within the original budget.
   
   ### Result
   
   Up to 120s of dilated wall-clock headroom on CI for the ActorSelection 
variant; still fails fast on genuine deadlocks.
   
   ### Tests
   
   - `sbt "remote / Test / compile"` — passes
   - CI Check / Test will exercise the artery variants
   
   ### References
   
   Refs #3041 (previous timeout widening from 10s to 30s)


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to