| Summary: | org.eclipse.rap.rwt.cluster.test.ServerPush_Test.testServerPushRequestDoesNotKeepSessionAlive is a toggling Test | ||
|---|---|---|---|
| Product: | [RT] RAP | Reporter: | Moritz Aleithe <moritz.aleithe> |
| Component: | RWT | Assignee: | Project Inbox <rap-inbox> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | ivan |
| Version: | unspecified | ||
| Target Milestone: | 3.7 M3 | ||
| Hardware: | PC | ||
| OS: | Windows 10 | ||
| See Also: |
https://git.eclipse.org/r/132505 https://git.eclipse.org/c/rap/org.eclipse.rap.git/commit/?id=ff2de823664298c9795832a79737c27145f21ffc |
||
| Whiteboard: | |||
|
Description
Moritz Aleithe
This is a known, very old issue... The test relies on Jetty session invalidation (scavenger to trigger). There is a timeout in the test to wait for Jetty. In some cases (server load) Jetty session invalidation takes a lot of time, which makes the timeout in the test not enough. We didn't find a better way to write this test. Hmm.. how about some "active waiting" to check whether Jetty closed in the maximum acceptable Time? This way, in the good case the tests are not blocked, and in the "worst case" of Jetty taking longer the test does not fail unneccessarily? Do you know how long Jetty tends to take? Could be 30 seconds or more... If you have a good fresh suggestion I'm open to hear it. As far as I see it, the Test waits for 3000 ms. This Is taken by the Constant SESSION_SWEEP_INTERVAL. I Could not find any other use of this Constant, so i wonder how that values was determined to be the optimal time for the test to wait? New Gerrit change created: https://git.eclipse.org/r/132505 Gerrit change https://git.eclipse.org/r/132505 was merged to [master]. Commit: http://git.eclipse.org/c/rap/org.eclipse.rap.git/commit/?id=ff2de823664298c9795832a79737c27145f21ffc |