Nicklas Bondesson wrote:
> Hi,
>
> I have posted before to the list regarding this problem we're having.
>
> The NDBD nodes crashes from time to time and spits the following message in
> the logs:
>
> Date/Time: Wednesday 25 May 2005 - 18:38:03 Type of error: error
> Message: Max number of open files exceeded Fault ID: 2806 Problem data:
> Object of reference: Ndbfs::createAsyncFile
> ProgramName: ndbd
> ProcessID: 8742
> TraceFile: /var/lib/mysql-cluster/ndb_3_trace.log.3
> Version 4.1.12
> ***EOM***
>
> This is a three server cluster setup. One management node, two storage nodes
> and three API nodes. All nodes are running the latest 4.1.11 MySQL source
> distribution. OS is Linux Debian 3.0 Woody with a 2.4.30 kernel.
>
> I have tried to raise the open files limit (ulimit -n shows: 10216 which I
> think should be more than enough).
>
> It would have been very nice if someone on the list could try to help me
> narrow down the problem.
>
>
> Thanks in advance (once again).
>
> Nicklas Bondesson
>
Hi Nicklas,
This is currently a known problem. Meaning that we know what causes it, but don't how to
fix it :-(
But, therefor I added a work around parameter, "MaxNoOfOpenFiles" (default is 40).
(I added it 2005-03-10)
Try increasing that to a big number say 100.
It would also be nice if you could mail the trace file, so that I can verify that it's the
know problem.
/Jonas
--
Jonas Oreland, Software Engineer
MySQL AB, www.mysql.com
