[
https://issues.apache.org/jira/browse/HCATALOG-616?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13575042#comment-13575042
]
Bill Graham commented on HCATALOG-616:
--------------------------------------
The logging issue comes into play in the case where you find there are bad
records beyond the threshold but you don't know just how bad it is. In that
case it could be useful to set a very high threshold and run the job just to
find the total good/bad counts. We need to make sure that's possible without
risk of over-logging.
> HCatRecordReader.InputErrorTracker should increment counters
> ------------------------------------------------------------
>
> Key: HCATALOG-616
> URL: https://issues.apache.org/jira/browse/HCATALOG-616
> Project: HCatalog
> Issue Type: Bug
> Reporter: Bill Graham
>
> It would be convenient if this class incremented counters to indicate good
> and bad records found.
> Also logging a warning for ever bad record seems risky w.r.t. excessive logs
> potentially filling up local disk. The frequency of the log messages should
> be configurable.
--
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
