[ 
https://issues.apache.org/jira/browse/SPARK-7621?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14542998#comment-14542998
 ] 

Jeremy A. Lucas edited comment on SPARK-7621 at 5/14/15 12:48 AM:
------------------------------------------------------------------

Attached is a small patch to simply call {{reportError}} instead of 
{{logError}} within the MessageHandler implementations of both the 
KafkaReceiver and the ReliableKafkaReceiver (note that the patch was generated 
for {{v1.3.1-rc3}}, but should be trivial to apply to other releases).


was (Author: jerluc):
Attached is a small patch to simply call {{reportError}} instead of 
{{logError}} within the MessageHandler implementations of both the 
KafkaReceiver and the ReliableKafkaReceiver.

> Report KafkaReceiver MessageHandler errors so StreamingListeners can take 
> action
> --------------------------------------------------------------------------------
>
>                 Key: SPARK-7621
>                 URL: https://issues.apache.org/jira/browse/SPARK-7621
>             Project: Spark
>          Issue Type: Improvement
>          Components: Streaming
>    Affects Versions: 1.3.0, 1.3.1
>            Reporter: Jeremy A. Lucas
>             Fix For: 1.3.1
>
>         Attachments: SPARK-7621.patch
>
>   Original Estimate: 24h
>  Remaining Estimate: 24h
>
> Currently, when a MessageHandler (for any of the Kafka Receiver 
> implementations) encounters an error handling a message, the error is only 
> logged with:
> {code:none}
> case e: Exception => logError("Error handling message", e)
> {code}
> It would be _incredibly_ useful to be able to notify any registered 
> StreamingListener of this receiver error (especially since this 
> {{try...catch}} block masks more fatal Kafka connection exceptions).



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

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org

Reply via email to