MySQL Forums
Forum List  »  NDB clusters

NDB seemingly losing data
Posted by: susan_h35
Date: August 14, 2006 11:46PM

We have been running MySQL 4.1.12 NDB storage engine for about a year with all MySQL processes (server and ndb nodes) running on the same Linux computer. Once in a while, we get strange ndbd hanging (not responding) with an error:

Can't lock file (errno: 4009)

but without ndbd crashing outright (we can see it using 'ps' command). In previous times, we could just manually kill these processes using 'kill -9' and restart without any data loss.

However, on Aug12 we got another errno: 4009 with similar symptoms but this time we seemed to have lost some data in our database! In our ndb_2_error.log file:

Date/Time: Saturday 12 August 2006 - 06:55:20
Type of error: error
Message: System error
Fault ID: 2303
Problem data: Node 2 killed this node because GCP stop was detected
Object of reference: NDBCNTR (Line: 193) 0x00000008
ProgramName: /usr/sbin/ndbd
ProcessID: 14803
TraceFile: /var/lib/mysql-cluster/ndb_2_trace.log.12
Version 4.1.12

We have also noticed that the regular local checkpoints were missing after July 9, 23:00. And then we get that 'GCP' error as above on Aug12.

Three questions:

1.) What could have happened to the regular checkpoints after July 9?
2.) What is a GCP? Is the lack of checkpoints for more than a month the cause of the 'GCP' error?
3.) Is losing data something we need to worry about with NDB?

Any suggestions or hints would be greatly appreciated.
Susan Henderson

Options: ReplyQuote


Subject
Views
Written By
Posted
NDB seemingly losing data
1457
August 14, 2006 11:46PM
981
August 15, 2006 08:13AM


Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.