Bug #36246 Incorrect handling of TC-TakeOver in cascading master failure

• Submitted: 22 Apr 2008 11:10
• Modified: 31 May 2008 12:47
• Reporter: Jonas Oreland
• Status: Closed
• Category: Server: Cluster
• Severity: S3 (Non-critical)
• OS: Any
• Assigned to: Jonas Oreland
• Triage: D1 (Critical) / R1 (None/Negligible) / E1 (None/Negligible)

[22 Apr 2008 11:10] Jonas Oreland

Description:
When a node X dies,
the master Y will start take-over of that nodes transactions (TC-TakeOver)

If the master dies during this take-over,
the new master Z will currently *not* complete the take-over of X

This can lead to stale operations on datanodes,
with pain and misery as a consequence

How to repeat:
will write new test prg

Suggested fix:
.

[25 Apr 2008 9:54] Jonas Oreland

pushed to 51-ndb, telco* and drop6
(50-ndb was locked for unknown reason)

[20 May 2008 11:34] Jon Stephens

Documented in the 5.1.24-ndb-6.3.14 changelog as follows:

        Under certain rare circumstances, the failure of the new master node
        while attempting a node takeover would cause takeover errors to repeat
        without being resolved.

Left Patch Queued status pending further merges.

[31 May 2008 12:47] Jon Stephens

Closed per yesterday's discussion with Jonas.

[13 Dec 2008 0:29] Bugs System

Pushed into 6.0.6-alpha  (revid:[e-mail]-20080423140838-48946)
(version source revid:[e-mail]-20080808094047-4e1yiarqa2t3opg3) (pib:5)