[ 
https://issues.apache.org/jira/browse/TINKERPOP-2144?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16752788#comment-16752788
 ] 

ASF GitHub Bot commented on TINKERPOP-2144:
-------------------------------------------

spmallette commented on pull request #1049: TINKERPOP-2144 Better handle 
Authenticator failures
URL: https://github.com/apache/tinkerpop/pull/1049
 
 
   https://issues.apache.org/jira/browse/TINKERPOP-2144
   
   Failures occurred on the first round of a SASL request when the negotiator 
was being created and if an exception was tossed there neither the server nor 
driver handled the situation well. By explicitly capturing and handling it, we 
can now properly assert results in kerberos tests without having to try to 
assert server logs which wasn't always dependable for some reason. This change 
should prevent random failures around kerberos tests which has long been a 
problem.
   
   All tests pass with `docker/build.sh -t -n -i`
   
   VOTE +1
 
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


> Better handle Authenticator instance failures
> ---------------------------------------------
>
>                 Key: TINKERPOP-2144
>                 URL: https://issues.apache.org/jira/browse/TINKERPOP-2144
>             Project: TinkerPop
>          Issue Type: Improvement
>          Components: server
>    Affects Versions: 3.3.5
>            Reporter: stephen mallette
>            Assignee: stephen mallette
>            Priority: Major
>
> There are handful of kerberos tests that randomly fail (depending on the 
> environment but mostly in docker for me). Investigation revealed that 
> {{Authenticator}} instances on the first SASL round of a request were not 
> trapping exceptions and were just firing pipeline errors which could hang a 
> client until timeout. By fixing that problem, we no longer need to rely on 
> the assertion of server logs for kerberos tests which should harden them from 
> random failure.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to