MySQL Bugs: #51389: Node-failure during online upgrade from 6.2 to 6.3

Bug #51389	Node-failure during online upgrade from 6.2 to 6.3
Submitted:	22 Feb 2010 13:55	Modified:	23 Feb 2010 10:25
Reporter:	Jonas Oreland	Email Updates:
Status:	Closed	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S3 (Non-critical)
Version:	mysql-5.1-telco-6.3	OS:	Any
Assigned to:	Jonas Oreland	CPU Architecture:	Any

Description:
In 6.3, support for read-before-update-removal was added.
This introduced a change in dbtc/dblqh protocol so that
commit ack markers was added on each operation until
on LQH replied.

This code was not made upgrade-safe, so that a 6.3 node
could crash a 6.2 by issuing several commit ack markers
causing DBLQH to crash with internal error.

How to repeat:
run 6.2 to 6.3 upgrade with sufficiently big upgrade traffic
using transaction with more than 1 DML(insert/update/delete) per transaction.

Suggested fix:
Make 6.3 DBTC aware of old 6.2-nodes in cluster.

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/101079

3097 Jonas Oreland	2010-02-22
      ndb - bug#51389 - fix commit ack marker handling wrt 62 to 63 upgrade

pushed to 6.3.32

Note: this bug fix also includes a fix for a crash in DBDIH::handle_invalid_lcp_no

Documented bugfix in the NDB-6.3.32 changelog as follows:

        During an online upgrade from MySQL Cluster NDB 6.2 to MySQL
        Cluster NDB 6.3, a sufficiently large amount of traffic with
        more than 1 DML operation per transaction could lead.an NDB 6.3
        data node to crash an NDB 6.2 data node with an internal error
        in the DBLOQH kernel block.

Closed.