MySQL Bugs: #47852: Crash after upgrade to 7.0.7

Bug #47852	Crash after upgrade to 7.0.7
Submitted:	6 Oct 2009 8:38	Modified:	8 Oct 2009 9:42
Reporter:	Daniel Herlitz	Email Updates:
Status:	Not a Bug	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S2 (Serious)
Version:	7.0.7	OS:	Linux
Assigned to:	Jonas Oreland	CPU Architecture:	Any

Description:
A couple of weeks ago we upgraded one of our systems from 7.0.6 to 7.0.7. Since this is a staging environment we frequently do full database drops + recreates from SQL. Usually no problems. (I mention this in case it might be of any relevance). 

This morning someone decided to reboot the machines and now we can't get NDB up and running again:

2009-10-06 09:58:36 [ndbd] ALERT    -- Node 2: Forced node shutdown completed. Occured during startphase 4. Caused by error 2341: 'Internal program error (failed ndbrequire)(Internal error, programming error or missing error message, please report a bug). Temporary error, restart node'.

Since I noticed similar bug reports regarding ndbmtd I think I should point out that we are running ndbd.

How to repeat:
Don't know

Suggested fix:
Don't know

Config.ini

Attachment: config.ini (application/octet-stream, text), 631 bytes.

ndb_2_error.log

Attachment: ndb_2_error.log (application/octet-stream, text), 1.04 KiB.

ndb_2_trace.log.2

Attachment: ndb_2_trace.log.2.gz (application/x-gzip, text), 37.40 KiB.

Daniel,

Could you please send the output from the ndb_error_reporter?

/Gustaf

FYI: The datanode crashes as head and tail meets in redo-log.
     This is not related to 7.0.7 per se.
     Increasing size of your redo-log
       (using either NoOfFragmentLogFiles or FragmentLogFileSize) should fix it.

Note: Changing size of redo-log, can only be done using either 
      initial-node-restart, or initial cluster start.

error report

Attachment: ndb_error_report_20091006153207.tar.bz2 (application/octet-stream, text), 231.10 KiB.

Output from ndb_error_reporter is attached. However, on the third attempt NDB acutally started for us, so the information might not be of any use to you. But thank you for the tip. Will look at increasing the redo log.

given that this is not a 7.0.7 bug
and it's more of a configuration error
i'll close this as not a bug.