Bug #58383 Unable to set inifinite TimeBetweenEpochsTimeout
Submitted: 22 Nov 2010 14:23 Modified: 24 Nov 2010 20:02
Reporter: Jonas Oreland Email Updates:
Status: Closed Impact on me:
None 
Category:MySQL Cluster: Cluster (NDB) storage engine Severity:S3 (Non-critical)
Version: OS:Any
Assigned to: Jonas Oreland CPU Architecture:Any

[22 Nov 2010 14:23] Jonas Oreland 
Description:
The values for TimeBetweenEpochsTimeout are currently limited to 32 seconds
this can in some convoluted cases be too little.

This patch
1) increase the max to 256s
2) add the possibility to completely disable the GCP monitor
   (which triggers GCP stop in GCP_SAVE or GCP_COMMIT takes too long)
   by setting TimeBetweenEpochsTimeout=0
3) Adds extra information in cluster log about currenly in use values
  and warnings every time gcp-save takes more than 1 minute,
  and gcp-commit takes more than 10s

How to repeat:
.

Suggested fix:
.
[22 Nov 2010 14:24] 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/124638

4013 jonas oreland	2010-11-22
      ndb - bug#58383 - allow setting unlimited gcp lags
[22 Nov 2010 14:36] Bugs System 
Pushed into mysql-5.1-telco-7.0 5.1.51-ndb-7.0.21 (revid:jonas@mysql.com-20101122142712-kopktgbn4ovwadwq) (version source revid:jonas@mysql.com-20101122142712-kopktgbn4ovwadwq) (merge vers: 5.1.51-ndb-7.0.21) (pib:21)
[23 Nov 2010 11:40] Jonas Oreland 
pushed to 7.0.21 and 7.1.10
[24 Nov 2010 20:02] Jon Stephens 
Documented in the NDB-7.0.21 and 7.0.10 changelogs as follows:

        The following changes have been made with regard to the
        TimeBetweenEpochsTimeout data node configuration parameter:

            The maximum possible value for this parameter has been
            increased from 32000 milliseconds to 256000 milliseconds.

            Setting this parameter to zero now disables the GCP monitor;
            this has the effect of disabling GCP stops.

            The current value of this parameter and a warning are
            written to the cluster log whenever a GCP save takes longer
            than 1 minute or a GCP save takes longer than 10 seconds.

Also updated parameter description in documentation.

Closed.