Bug #48143 ndb_mgmd - crash in CONFIG_CHECK_REF
Submitted: 19 Oct 11:22 Modified: 2 Nov 16:39
Reporter: Magnus Blaudd
Status: Closed
Category:Server: Cluster Severity:S3 (Non-critical)
Version:7.0.9 OS:Any
Assigned to: Magnus Blaudd Target Version:

[19 Oct 11:22] Magnus Blaudd 
Description:
The CONFIG_CHECK that each ndb_mgmd run to check that all connected ndbd_mgmd's have the
same configuration may fail with an assert if a configuration change occurs while the
check is ongoing. A configuration change is allowed to take place during the check and
the check should simply be scheduled to be run again at a later time when the
configuration change has completed.

How to repeat:
testMgmd

Suggested fix:
Drop the CONFIG_CHECK_REF and thus cause the check to be run again later.
[19 Oct 14:07] Magnus Blaudd 
Pushed to 7.0
[2 Nov 16:39] Jon Stephens 
Documented bugfix in the NDB-7.0.10 changelog as follows:

        The configuration check that each management server runs to
        verify that all connected ndbd_mgmd processes have the same
        configuration could fail when a configuration change took place
        while this check was in progress. Now in such cases, the
        configuration check is rescheduled for a later time, after the
        change is complete.

Closed.