Bug #8549 transactions aborted in case of node failure
Submitted: 16 Feb 2005 15:41 Modified: 10 Apr 2005 17:43
Reporter: Jörg Nowak Email Updates:
Status: No Feedback Impact on me:
None 
Category:MySQL Cluster: Cluster (NDB) storage engine Severity:S2 (Serious)
Version:4.1.9 OS:Linux (Suse 9 64 bit)
Assigned to: Assigned Account CPU Architecture:Any

[16 Feb 2005 15:41] Jörg Nowak 
Description:
In case of node failure the cluster seems to fail in recognizing the node failure immediatedly and the queries are send to the failed node, resulting in the abort of transaction.

From the client side we got the following errors:
Got Temporary error: 4010. Transaction aborted due to node failure.
Got Temporary error: 4119. Simple/dirty read failed due node failure.

How to repeat:
make a continious read/write operations randomly in different nodes and crash one node.

One the client side you will get this error messages.
[16 Feb 2005 18:44] Tomas Ulin 
transactions may be aborted due to nodefailure.

these transactions must be "retried" from the application.

in the future we may implement retries "under the hood" in certain cases, but this will not be in 4.1 or 5.0.
[14 Apr 2005 23:00] Bugs System 
No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".