MySQL Bugs: #12211: Error Attempting Cluster Backup

Bug #12211	Error Attempting Cluster Backup
Submitted:	27 Jul 2005 15:45	Modified:	20 Oct 2005 16:26
Reporter:	Stewart Burnett	Email Updates:
Status:	No Feedback	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S3 (Non-critical)
Version:	4.1.11-max	OS:	Linux (SLES 9.1 (On Open Power))
Assigned to:	Assigned Account	CPU Architecture:	Any

Description:
We are getting the following when attempting to perform a Cluster Backup

Connected to Management Server at: 172.23.77.231:1186
Waiting for completed, this may take several minutes
Node 2: Backup request from 1 failed to start. Error: 1302
Start of backup failed
*  3001: Could not start backup
*        Out of backup record: Permanent error: Internal error

ndb_1_cluster.log contains

2005-07-27 16:37:21 [MgmSrvr] ALERT    -- Node 2: Backup request from 1 failed to start. Error: 1302

Config.ini is

[NDBD DEFAULT]
NoOfReplicas=1

DataMemory=1600Mb
IndexMemory=200Mb

MaxNoOfAttributes=2000

MaxNoOfTables=256
MaxNoOfUniqueHashIndexes=256
MaxNoOfOrderedIndexes=256

FileSystemPath=/mysqlcluster
BackupDataDir=/mysqlbackup

MaxNoOfConcurrentOperations=500000

TransactionDeadlockDetectionTimeout=2400

[MYSQLD DEFAULT]
[NDB_MGMD DEFAULT]
[TCP DEFAULT]

SendBufferMemory=512KB

# Managment Server
[NDB_MGMD]
Id=1
HostName=172.23.77.231
DataDir=/usr/local/mysql/cluster

# Storage Engines
[NDBD]
Id=2
HostName=172.23.77.232
DataDir=/usr/local/mysql/cluster

[NDBD]
Id=3
HostName=172.23.77.232
DataDir=/usr/local/mysql/cluster

[NDBD]
Id=4
HostName=172.23.77.232
DataDir=/usr/local/mysql/cluster

[NDBD]
Id=5
HostName=172.23.77.232
DataDir=/usr/local/mysql/cluster

[MYSQLD]
[MYSQLD]

No output in ndbd log files.

How to repeat:
Not sure, have not had a chance to re-start the cluster to try and fix the problem.

However,the backup was working correctly. Only thing I'm aware of is that the cluster was re-started with an additional storage node (after advice recieved about 2^n nodes) and the backup sequence reverted back to 1. The backup would then have encountered previous backup diredtories files etc after a few backups.

Re-starting the ndbd processes has cleared the problem.

Hi,

1302 can occur in the following situation:
* A started backup was interupted by node shutdown
  In _some_ situations, one alive node fails to abort correctly 
  with 1302 as a result on next backup.
  In this case this node can be restarted.

* Trying to start a backup _imediatly_ after a backup has finished.
  The release of resources is done in a lazy manner after the backup has completed.

* Trying to start a backup while a backup is running.

Does any of these fit your problem?
/Jonas

ps.
there has be some bug fixes wrt backup in the latest 4.1
(regarding simultanious  ddl)
ds.

I'm getting the error when running two backups close together. Your comments explain why re-starting the storage nodes helps.

Does this mean that I can close the bug?

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".