MySQL Bugs: #47188: Data node crash on table alter to NDB

Bug #47188	Data node crash on table alter to NDB
Submitted:	8 Sep 2009 13:36	Modified:	14 Oct 2009 9:09
Reporter:	mike grafton	Email Updates:
Status:	No Feedback	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S3 (Non-critical)
Version:	mysql-5.1-telco-7.0	OS:	Linux (RHEL 2.6.18-128.el5)
Assigned to:	Assigned Account	CPU Architecture:	Any
Tags:	5.1.34-ndb-7.0.6, Forced node shutdown completed. Caused by error 2815

Description:
Getting the following error when altering a table from InnoDB to NDB:

Forced node shutdown completed. Caused by error 2815: 'File not found(Ndbd file system inconsistency error, please report a bug). Ndbd file system error, restart node initial'.

This causes ndbd to crash on data nodes (area105 and area106) but everything else stays up and running.

I have been able to alter several small (<10 row) tables w/o problems. This one has several million rows so I would guess the problem is volume-related.

Configuration has 3 RH Linux servers;
area50 - management node - 8 GB RAM
area105 - data and sql nodes - 32 GB RAM
area106 - data and sql nodes - 48 GB RAM 

How to repeat:
Restart data nodes and issue alter command.

Mike,
It seems the tar ball generated by ndb_error_reporter wasn't uploaded correctly. Can you try again?

Also, are these NDB tables using the Disk-Based feature?

ndb error report added.

not using disk-based.

thanks.

Also, I am not able to re-produce this error as before.  I did get it one time earlier this week but not as reliably as I was getting it last week.

Hi,

It looks like you changed the value of NoOfFragmentLogfiles wo/ having
restarted the nodes (rolling) initial.

/Jonas

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".