You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@accumulo.apache.org by "Mike Drob (JIRA)" <ji...@apache.org> on 2015/06/22 23:54:00 UTC

[jira] [Commented] (ACCUMULO-3912) Random walk sets random config on default namespace

    [ https://issues.apache.org/jira/browse/ACCUMULO-3912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14596726#comment-14596726 ] 

Mike Drob commented on ACCUMULO-3912:
-------------------------------------

As a broader issue, would it make sense to place random-walk and continuous-ingest and whatever else to "test" table namespace?

> Random walk sets random config on default namespace
> ---------------------------------------------------
>
>                 Key: ACCUMULO-3912
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-3912
>             Project: Accumulo
>          Issue Type: Bug
>            Reporter: Keith Turner
>
> While running random walk test for 1.6.3, the following failure happened.
> {noformat}
> Caused by: java.lang.Exception: Props not equals ST_index_worker4_3842_1435004444639 ST_index_worker4_3842_1435004444639_tmp
>         at org.apache.accumulo.test.randomwalk.shard.ExportIndex.visit(ExportIndex.java:106)
> {noformat}
> I looked at the two tables config in the shell and found the following property differed. 
> {noformat}
> default    | table.compaction.minor.idle ............... | 5m
> default    | table.compaction.minor.logs.threshold ..... | 3
> {noformat}
> The _tmp table had the following setting.
> {noformat}
> default    | table.compaction.minor.idle ............... | 5m
> table      |    @override .............................. | 178299085
> default    | table.compaction.minor.logs.threshold ..... | 3
> {noformat}
> I grepped all the random walkers logs and found the following :
> {noformat}
> [11] 21:18:42 [SUCCESS] worker0
> 22 20:33:57,007 [concurrent.Config] DEBUG: Setting table.compaction.minor.idle on namespace  to 178299085
> {noformat}
> So {{concurrent.Config}} updated props on a random namepace and caused the shard test to fail, it should not do this.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)