Bug #39453 Replication from Cluster to MySQL causes PK violation on ndb_apply_status

• Submitted: 15 Sep 2008 13:06
• Modified: 9 May 2009 13:02
• Reporter: Oli Sennhauser
• Status: No Feedback
• Category: MySQL Cluster: Replication
• Severity: S2 (Serious)
• Version: 5.1.27-ndb-6.3.17
• OS: Any
• CPU Architecture: Any

[15 Sep 2008 13:06] Oli Sennhauser

Description:
We have a MySQL Cluster where we store our OLTP stuff in. We want to replicate the whole traffic to a MySQL database to store all our OLAP stuff in.

Replication breaks continuously with PK violations on ndb_apply_status.

This was working in earlier versions. And we recommend it sometimes to customers...

How to repeat:
Create Cluster with 1 Master. Attach slave. Insert some rows into Cluster. -> Replication breaks.

Suggested fix:
No idea.

There are several workarounds:
* filtering on that table (no more channel failover possible)
* alter ndb_apply_status to blackhole (no more channel failover possible)
* drop pk on ndb_apply_status (we have to clean-up the table ourself)

[9 Apr 2009 13:02] Jonathan Miller

Please let us know if this happens on 6.3.20 -> GA release

Also need how to reproduce and link to support case.

[9 May 2009 23:00] Bugs System

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".