Hi Pekka,
I set in my.cnf:
ndb-log-bin = 0
ndb-log-binlog-index = 0
and turned off regular bin-logging. Restarted mysqld. Re-ran benchmark
and same thing. Seems to happen when the benchmark hits 64 threads.
Interestingly, if I run the "update_index" bench (ie: UPDATE table SET
indexedColum=indexedColumn+1 WHERE PKcolumn=<rand>), that one runs just
fine all the way up to 256 threads.
But if I try the update_nonindex, (ie: UPDATE table SET
nonIdxColum='<random 100 char string>' WHERE PKcolumn=<rand>), that is
the one that fails right away on the 64/128/256 thread bench.
I'm going to try and modify sysbench to change that from a 100 char rand
string to 50 and see if that makes a difference.
-Matthew
> -----Original Message-----
> From: Pekka.Nousiainen@stripped [mailto:Pekka.Nousiainen@stripped]
> Sent: Tuesday, November 10, 2009 11:57 AM
> To: cluster@stripped
> Subject: Re: Send Buffers overloaded in NDB kernel
>
> On 091110, Boehm, Matthew wrote:
>
> > I find it difficult to understand how 256M of buffer can be used up
> when
> > all 3 machines (2 ndbd, 1 mysqld) are on the same gigE switch.
>
> The "Send Buffers overloaded" error was added specifically for
> overloaded binlogging to mysqld. The mysqld receives events
> (data changes) from all DB nodes. Either the connections from DB
> nodes,
> or the mysqld, or an attached replication channel cannot keep up.
>
> In this case it would be the first one (DB events). Or else you
> have found a new case for this error. Not sure if binlogging can
> be disabled, others here will know.
>
> Setting the buffer to 256M or higher does not help. It cannot
> keep up, so it will soon be full.
>
> --
> Pekka Nousiainen, Software Engineer, Sun Microsystems / MySQL
>
> --
> MySQL Cluster Mailing List
> For list archives: http://lists.mysql.com/cluster
> To unsubscribe:
> http://lists.mysql.com/cluster?unsub=1
