Bug #21953 Master threads stays, slaves hangs on restart
Submitted: 31 Aug 2006 18:47 Modified: 6 Nov 2006 13:03
Reporter: Geert Vanderkelen Email Updates:
Status: Can't repeat Impact on me:
Category:MySQL Cluster: Replication Severity:S2 (Serious)
Version:5.1.2-drop5p14 OS:
Assigned to: Mats Kindahl CPU Architecture:Any
Tags: replication

[31 Aug 2006 18:47] Geert Vanderkelen
Having a replication going using Row-Based Replication between Clusters:

When the slave is killed like using SIGKILL (kill -9 <pid>) the master thread stays active on the master. If you then restart the slave, commands like 'SHOW SLAVE STATUS' is hanging. The Master Thread is not getting the new connection.

This does happen in 5.1.2-drop5p14, but not using 5.1.11.

How to repeat:
See above.