[
https://issues.apache.org/jira/browse/GEODE-8644?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17454924#comment-17454924
]
Xiaojian Zhou commented on GEODE-8644:
--------------------------------------
What raised my attention is:
1) In all failed tests, there're
{code:java}
[vm5] [info 2021/12/04 03:37:28.229 UTC <poolTimer-ln-4> tid=0x69] Socket
receive buffer size is 212992 instead of the requested 524288.
[vm5] [info 2021/12/04 03:37:28.230 UTC <poolTimer-ln-4> tid=0x69] Socket send
buffer size is 212992 instead of the requested 524288.
[vm4] [info 2021/12/04 03:37:32.891 UTC <poolTimer-ln-5> tid=0x6c] Socket
receive buffer size is 212992 instead of the requested 524288.
[vm4] [info 2021/12/04 03:37:32.891 UTC <poolTimer-ln-5> tid=0x6c] Socket send
buffer size is 212992 instead of the requested 524288.
{code}
In the passed tests, there's no such message.
2) There're much more locator error message:
{code:java}
[locator] [info 2021/12/04 03:37:42.421 UTC <WAN Locator Discovery Thread1>
tid=0x46] Failed to connect to localhost/127.0.0.1:0
{code}
In passed tests, there're about less than 5 of these messages, but in failed
tests, there're 15-20 of these message.
> SerialGatewaySenderQueueDUnitTest.unprocessedTokensMapShouldDrainCompletely()
> intermittently fails when queues drain too slowly
> -------------------------------------------------------------------------------------------------------------------------------
>
> Key: GEODE-8644
> URL: https://issues.apache.org/jira/browse/GEODE-8644
> Project: Geode
> Issue Type: Bug
> Affects Versions: 1.15.0
> Reporter: Benjamin P Ross
> Assignee: Mark Hanson
> Priority: Major
> Labels: GeodeOperationAPI, needsTriage, pull-request-available
>
> Currently the test
> SerialGatewaySenderQueueDUnitTest.unprocessedTokensMapShouldDrainCompletely()
> relies on a 2 second delay to allow for queues to finish draining after
> finishing the put operation. If queues take longer than 2 seconds to drain
> the test will fail. We should change the test to wait for the queues to be
> empty with a long timeout in case the queues never fully drain.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
