[
https://issues.apache.org/jira/browse/ACCUMULO-2194?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13931757#comment-13931757
]
Bill Havanki commented on ACCUMULO-2194:
----------------------------------------
Got this a second time, but different error:
{noformat}
Caused by: org.apache.accumulo.core.client.AccumuloException: User didn't exist
when they should (or
worse- password mismatch)
at
org.apache.accumulo.server.test.randomwalk.security.TableOp.visit(TableOp.java:62)
at
org.apache.accumulo.server.test.randomwalk.Module.visit(Module.java:251)
{noformat}
The common thread in both cases is that the table user is found to exist when
the user was in fact dropped at the end of the prior run. I'm pretty sure,
based on similar issues before, that it's just that ZK hasn't had time enough
to completely absorb the change. (When starting a new security test, the system
user is created but the table user is not, so this is not a problem for the
system user account.) This time here, the time between dropping the user and
the error is just under one second.
> Back-to-back Security randomwalk tests fail
> -------------------------------------------
>
> Key: ACCUMULO-2194
> URL: https://issues.apache.org/jira/browse/ACCUMULO-2194
> Project: Accumulo
> Issue Type: Bug
> Components: test
> Affects Versions: 1.4.4
> Reporter: Bill Havanki
> Assignee: Bill Havanki
> Labels: randomwalk, test
> Attachments: Security_failure_log.txt
>
>
> When the Security randomwalk test is run twice in a row by a higher-level
> test (e.g., ShortClean), the second Security test fails relatively quickly. A
> sample log is attached.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
