cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Michael Yeh (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-4856) Thrift daemon is crashing the server
Date Thu, 25 Oct 2012 19:49:12 GMT


Michael Yeh commented on CASSANDRA-4856:

Setting rpc_max_threads worked.  I'm seeing: 

WARN 12:35:13,255 Maximum number of clients 1000 reached

instead of the server crashing.  Would it be better to set an artificially high limit vs unlimited
number of threads as the default?  IMHO, I would prefer to deal with some tuning on the server
than one that crashes because of misbehaving clients which may or may not be under your control.
 Thanks for the tip.
> Thrift daemon is crashing the server
> ------------------------------------
>                 Key: CASSANDRA-4856
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>    Affects Versions: 1.1.6
>         Environment: MacOS 10.8.2
>            Reporter: Michael Yeh
> I'm new to cassandra and was looking at using the kundera thrift client.  I was able
to get the sample code working without any issues.  After upgrading to mountain lion on my
mac, I didn't realize I had lost the /var/lib/cassandra directory.  When I tried to run the
previously working sample code, it would crash cassandra with the following stack trace:
> INFO 12:55:37,147 Listening for thrift clients...
> ERROR 12:55:43,545 Exception in thread Thread[Thread-3,5,main]
> java.lang.OutOfMemoryError: unable to create new native thread
> 	at java.lang.Thread.start0(Native Method)
> 	at java.lang.Thread.start(
> 	at java.util.concurrent.ThreadPoolExecutor.addIfUnderMaximumPoolSize(
> 	at java.util.concurrent.ThreadPoolExecutor.execute(
> 	at org.apache.cassandra.thrift.CustomTThreadPoolServer.serve(
> 	at org.apache.cassandra.thrift.CassandraDaemon$
> After realizing that I needed to recreate the keyspace and column families, everything
started to work again.  Although I'm no longer blocked with my issue, no clients should ever
be able to crash a server even if it is user error.

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:

View raw message