Bug #26310 Backup only on one data node

• Submitted: 13 Feb 2007 9:20
• Modified: 13 Mar 2009 8:42
• Reporter: Stewart Smith
• Status: Closed
• Category: MySQL Cluster: Cluster (NDB) storage engine
• Severity: S3 (Non-critical)
• OS: Any
• Assigned to: Stewart Smith
• CPU Architecture: Any

[13 Feb 2007 9:20] Stewart Smith

Description:
This one is straight from the forum:

http://forums.mysql.com/read.php?25,139178,139178

I think it may be something with checking success/fail of FSOPENREQ in BACKUP. Needs to be verified though.

How to repeat:
I am testing a two data node cluster for 5.1, because we are going to go into production. I issue the START BACKUP WAIT COMPLETED from the management console (10.11.11.5), which is also data node 1 and SQL node 1 (for testing purposes). It will say

ndb_mgm> start backup wait completed
Connected to Management Server at: 10.11.11.5:1186
Waiting for completed, this may take several minutes
Node 5: Backup 2 started from node 1
Node 5: Backup 2 started from node 1 completed
StartGCP: 73892 StopGCP: 73896
#Records: 1571385 #LogRecords: 0
Data: 132563392 bytes Log: 0 bytes

When it is completed, I go and look at BACKUP-2 and I see data from today, but when I go to the other data node (10.11.11.15) I don't see new data in the BACKUP-2 folder, its old. Since I only have one node group, with two data nodes in it, does it only perform the backup off of one snapshot of the data?

[root@db0 BACKUP-2]# ls -l
total 66160
-rw-r--r-- 1 root root 66624136 Feb 9 10:00 BACKUP-2-0.5.Data
-rw-r--r-- 1 root root 1041372 Feb 9 10:00 BACKUP-2.5.ctl
-rw-r--r-- 1 root root 44 Feb 9 10:00 BACKUP-2.5.log

[root@db0-1 BACKUP-2]# ls -l
total 65512
-rw-r--r-- 1 mysql mysql 65961680 Feb 8 23:06 BACKUP-2-0.6.Data
-rw-r--r-- 1 mysql mysql 1041372 Feb 8 23:06 BACKUP-2.6.ctl
-rw-r--r-- 1 mysql mysql 44 Feb 8 23:06 BACKUP-2.6.log

Suggested fix:
check if files exist, replace or error out