Bug #34216 Multi-update with ordered index can crash node during nf-handling
Submitted: 1 Feb 2008 9:32 Modified: 20 Feb 2008 21:05
Reporter: Jonas Oreland Email Updates:
Status: Closed Impact on me:
None 
Category:MySQL Cluster: Cluster (NDB) storage engine Severity:S3 (Non-critical)
Version:5.1* OS:Any
Assigned to: Jonas Oreland CPU Architecture:Any

[1 Feb 2008 9:32] Jonas Oreland
Description:
During TC-take-over (during NF)
commit messages can come out of order to TUP
that can make the OI-triggers to be run in incorrect order

How to repeat:
will write new test prg
[1 Feb 2008 12:37] Jonas Oreland
Patch

Attachment: pp (application/octet-stream, text), 11.41 KiB.

[6 Feb 2008 10:28] Bugs System
A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/41756

ChangeSet@1.2679, 2008-02-06 11:28:43+01:00, jonas@perch.ndb.mysql.com +6 -0
  ndb - bug#34216 
    testcases
[6 Feb 2008 12:05] Bugs System
A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/41766

ChangeSet@1.2680, 2008-02-06 13:06:05+01:00, jonas@perch.ndb.mysql.com +5 -0
  ndb - bug#34216
    During TC-take-over (NF) the new-TC builds up a new transaction state
    And commits operation according to this state.
    However, in the new state that is build, the operations does not have to be in same order, as "real" state
    In the multi-update-case, this means that operations can be commit in "incorrect" order
  
    i.e update A, delete A, insert A is normally commited in same order as prepared
        but can be committed in any order
  
    This patch changes TUP handling of these out-order commits, and previous implementation
      could confuse the TUX triggers
[6 Feb 2008 12:37] Jonas Oreland
pushed to 51-ndb, and telco 62,63,64
[12 Feb 2008 15:01] Jon Stephens
Documented bugfix in the 5.1.23-ndb-6.3.9 changelog as follows:

        Multi-table updates using ordered indexes during handling of
        node failures could cause other data nodes to fail.

Left in PQ status pending further merges.
[12 Feb 2008 16:09] Jon Stephens
Also documented for 5.1.23-ndb-6.2.12.
[20 Feb 2008 16:03] Bugs System
Pushed into 5.1.24-rc
[20 Feb 2008 16:04] Bugs System
Pushed into 6.0.5-alpha
[20 Feb 2008 21:05] Jon Stephens
Also documented for 5.1.24 and 6.0.5.
[20 Aug 2008 20:28] Bugs System
A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/52094

2653 Jonas Oreland	2008-08-20
      ndb - fix testNodeRestart -n Bug34216