[
https://issues.jboss.org/browse/RFPL-3692?page=com.atlassian.jira.plugin....
]
Jiří Štefek commented on RFPL-3692:
-----------------------------------
After some debugging:
* the push still works even in this "freezed state",
* there is an error in browser console every time the push is enabled/disabled:
{{no element found @ __richfaces_push:1}}
It seems that the Selenium will stuck, when this error happens while it is communicating
with the browser.
After adding some small wait time before the waiting method (after push is
enabled/disabled), the test seems to be stabilized.
metamer: fix push tests
-----------------------
Key: RFPL-3692
URL:
https://issues.jboss.org/browse/RFPL-3692
Project: RichFaces Planning
Issue Type: Task
Components: QE
Affects Versions: 4.5.1.Final
Environment: Firefox
Reporter: Jiří Štefek
Assignee: Jiří Štefek
Only on *Firefox*.
Sometimes, running these tests freezes the whole suite inside one test method.
Log from Jenkins (check the times):
{code}
06:26:31.869 [09:10:00] STARTED: a4jPush.TestTwoPushCDI.testPushEnable() { template=plain
}
09:11:17.998 Build timed out (after 180 minutes). Marking the build as aborted.
09:11:18.003 Terminating xvnc.
09:11:18.011 $ vncserver -kill :41
09:11:18.144 Killing Xvnc process ID 20637
09:11:18.281 [11:54:46] FAILURE: a4jPush.TestTwoPushCDI.testPushEnable() { template=plain
}
{code}
It could be also reproduced locally by e.g.:
{{mvn verify -Dbrowser=firefox -DexcludedGroups=extended,Future -Dtemplates=*
-Pwildfly-managed-8-2 -Dtest=TestTwoPush#testOnSubscribed}}
It is hard to reproduce it sometimes, so try to run it repeatedly.
It will stuck on this line
[*Graphene.waitModel().until().element(page.getRequestTimeElement()).text().not().equalTo(requestTime);*|https://github.com/richfaces/richfaces-qa/blob/master/metamer/ftest/src/test/java/org/richfaces/tests/metamer/ftest/a4jPush/TestTwoPush.java#L59].
--
This message was sent by Atlassian JIRA
(v6.3.11#6341)