MySQL Bugs: #39318: NDB data node crashes

Bug #39318	NDB data node crashes
Submitted:	8 Sep 2008 15:43	Modified:	10 Sep 2008 10:49
Reporter:	Matthew Robinson	Email Updates:
Status:	Closed	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S2 (Serious)
Version:	5.0.51	OS:	Linux (Debian Etch)
Assigned to:		CPU Architecture:	Any
Tags:	crash, data node

Description:
Since this morning, one data node has crashed 3 times with the following error:

Time: Monday 8 September 2008 - 16:08:27
Status: Temporary error, restart node
Message: Temporary on access to file (Internal error, programming error or missing error message, please report a bug)
Error: 2809
Error data: DBACC: File system open failed. OS errno: 5
Error object: DBACC (Line: 1819) 0x0000000e
Program: /usr/sbin/ndbd
Pid: 568
Trace: /var/lib/mysql-cluster/ndb_3_trace.log.7
Version: Version 5.0.51
***EOM***

I have done an initial start both times, which temporarily fixed the issue.

How to repeat:
I dont know how to repeat

Suggested fix:
I dont know how to fix

Hi Matthew,

can you attach the config file, cluster logs as produced by the management server, out files and full trace files? You can the command ndb_error_reporter to collect all those files.

Uploaded to debian FTP site as:

ndb_error_report_20080909091521.tar.bz2

Sorry - i meant mysql ftp site:

ftp://ftp.mysql.com/pub/mysql/upload/

It seems I was premature with telling you I had uploaded it to the FTP site - I had some problems with that. Anyway, it can be downloaded from our webserver:

https://www.fone-me.com/ndb_error_report_20080909091521.tar.bz2

Update - no crashes since the bug was reported. Our usage hasn't changed ( we've been running the same setup for a couple of months ).

Quite likely you have/had a file system / disk problem as OS error 5 is io error not related to cluster in the first place. 

I will close bug for now as "not a bug" but you are welcome to open it again if you have other findings.

Ok, thanks for looking.