I think we can check for existence via thrift, and assume that the table is 
valid and working. What do you think?

Sent from my iPhone

On Sep 17, 2012, at 9:28 AM, "Tim Williams (JIRA)" <[email protected]> wrote:

> 
>    [ 
> https://issues.apache.org/jira/browse/BLUR-22?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13457001#comment-13457001
>  ] 
> 
> Tim Williams commented on BLUR-22:
> ----------------------------------
> 
> Honestly I'm not sure how to go about that, but sure, if you say so:)
> 
> I'm not sure what the point of truth is here and if existence in HDFS is 
> enough.  The only thing it seems to be doing there is checking for table 
> existence... which it seems like we could just use the 
> fileSystem.exists(tablePath)? But then, I reckon the path could exists but 
> not be an actual index.  I reckon it depends how robust we should be with 
> checking table existence on an UPDATE. 
> 
> 
> 
>> Zookeeper not set in BlurTask job on updates
>> --------------------------------------------
>> 
>>                Key: BLUR-22
>>                URL: https://issues.apache.org/jira/browse/BLUR-22
>>            Project: Apache Blur
>>         Issue Type: Bug
>>           Reporter: Tim Williams
>> 
>> When INDEXING_TYPE.UPDATE with a BlurTask, we try to checkTable() on job 
>> configuration which tries to go to zookeeper, only _zooKeeper is never 
>> actually set. 
>> I think it's fair to ask that the "blur.zookeeper.connection" property is 
>> set on the incoming job Configuration - which would allow 
>> ZkUtils.newZooKeeper?
> 
> --
> This message is automatically generated by JIRA.
> If you think it was sent incorrectly, please contact your JIRA administrators
> For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to