accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Josh Elser (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-2549) ShellTest sometimes fails because of order of authentication strings returned
Date Tue, 25 Mar 2014 18:04:16 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-2549?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13946866#comment-13946866
] 

Josh Elser commented on ACCUMULO-2549:
--------------------------------------

That's interesting. I thought we sort auths before they get returned (maybe that's done by
the shell after the API call?). Would be good to figure out what the implementation is going
to return and make sure the test case adheres to that. Seems odd behavior to "sometimes" fail.

That said, patch looks good to fix it too ignoring what the actual getauths implementation
is doing.

> ShellTest sometimes fails because of order of authentication strings returned
> -----------------------------------------------------------------------------
>
>                 Key: ACCUMULO-2549
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-2549
>             Project: Accumulo
>          Issue Type: Bug
>          Components: client
>    Affects Versions: 1.5.1
>            Reporter: Michael Allen
>            Priority: Minor
>             Fix For: 1.5.2, 1.6.1
>
>         Attachments: ACCUMULO-2549-Fixes-ShellTest.patch
>
>
> {{org.apache.accumulo.core.util.shell.ShellTest}} sometimes fails for me when run from
the Maven command line.  Tracing down, I see that line 154 (1.5.1 version) contains the following
> {noformat}
>     exec("getauths", true, "y,z,x");
> {noformat}
> This test expects that the set of authorizations returns from the {{getauths}} command
always come in the particular order "y,z,x".  For whatever reason, on my machine, they came
out in a different order ("x,y,z" for the record).
> My proposed fix (patch shortly) would be to have the test look for those individual label
strings returned within the authorization string.  I think this is a more accurate and future
proof test than what's in there now.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message