MySQL Bugs: #67260: in ndb7.2 all data node was downed

Bug #67260	in ndb7.2 all data node was downed
Submitted:	16 Oct 2012 10:22	Modified:	18 Feb 2013 18:01
Reporter:	ws lee	Email Updates:
Status:	No Feedback	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S1 (Critical)
Version:	mysql5.5.27-ndb7.2.8	OS:	Solaris (SunOS 5.10)
Assigned to:	Assigned Account	CPU Architecture:	Any

Description:
I recently have upgraded mysql cluster from mysql5.1.51-ndb7.1.10 to mysql5.5.27-ndb7.2.8.

and  setted in config.ini
ndb_push_joindown=1

after a few days(today), all data node downed.

error message is below.

Time: Tuesday 16 October 2012 - 18:29:43
Status: Temporary error, restart node
Message: Internal program error (failed ndbrequire) (Internal error, programming error or missing error message, please report a bug)
Error: 2341
Error data: Message to big in sendSignal
Error object: 
Program: ndbmtd
Pid: 26317 thr: 0
Version: mysql-5.5.27 ndb-7.2.8
Trace: /var/lib/mysql-cluster/mysql5.5.27-ndb7.2.8/ndb_4_trace.log.1 [t1..t5]

i pay attention to below message.

Error data: Message to big in sendSignal

it seems that this error is connected ndb_join_pushdown=1.

How to repeat:
ndb_join_pushdown=1

run heavy query with big traffic.

Suggested fix:
It seems that this problem is connected ndb_join_pushdown=1

To. mysql staff

why not check this bug?
I have confidence "ndb_join_pushdown" new optoin cause critical problem.

To. mysql cluster staff

how do you think ndb_join_pushdown option's instability?

I experienced cluster down horrbly after using ndb_join_pushdown.

Hi,

Can you provide us a sample data set and query/ies to reproduce this issue locally? Or, if possible, please let us know how to recreate the data set locally.

I am assuming that the issue never showed up in previous version of cluster. Also, include configuration files (eg. config.ini, my.cnf), log files (or NDB error report), and any other info that might be helpful.

Thanks.

No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".