[
https://issues.apache.org/jira/browse/SAMZA-137?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13891429#comment-13891429
]
Jakob Homan commented on SAMZA-137:
-----------------------------------
You're referring to:
{noformat} Triggering a reconnect for kafka because connection failed:
java.lang.String cannot be cast to [B{noformat}?
I believe that's the whole message and is just Java's poor attempt at meaning
byte array. See
[here|http://grokbase.com/t/kafka/users/141qnrpaxh/error-in-kafka-producer] The
whole stack trace may be useful (though repetitive and has been a question of
usefulness here bfore).
> Fix how long warning messages get truncated in container logs
> -------------------------------------------------------------
>
> Key: SAMZA-137
> URL: https://issues.apache.org/jira/browse/SAMZA-137
> Project: Samza
> Issue Type: Bug
> Affects Versions: 0.7.0
> Reporter: Jonathan Poltak Samosir
> Priority: Minor
> Labels: log, logs
> Fix For: 0.7.0
>
>
> If there is a Java problem encountered that produces a long warning message,
> that warning message will be truncated in the given container's logs. This is
> an issue as it makes it painful to solve as issue that could otherwise be
> solved simply by reading the full Java warning message.
> An example of what I am talking about can be found in one of my logs, here
> (scroll down a few lines until you see the recurring truncated warning
> message):
> https://gist.github.com/poltak/8745919/raw/7990b0199fafde9bf6da00aa4b87b2e14fb0e717/gistfile1.txt
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
