BMR restore failed

[walker_yang]

Hello, everybody, i have a problme about BMR restore, detailed enviorment as follows:

NBU master(BMR master/BMR boot server): 7.0.1, SuSE 11 SP1 / NBU Client: 7.0.1, SuSE 10 SP2.

BMR backup is OK, when restoring BMR via network to the same client, restore operation failed. During the restore, it prompts inputing a new discovered configuration name, after that, it shows shell mode. Then i collect logs from /usr/openv/netbackup/logs/bmrrst/CLINET_NAME/xxx.log.

=================================================================================

' /proc/scsi/scsi

++ eval 'ScsiDiskLocs[0]="scsi0:00:00:00"' 'ScsiDiskNames[0]="sda"'
'ScsiDiskLocs[1]="scsi0:00:01:00"' 'ScsiDiskNames[1]="sdb"'
'ScsiDiskLocs[2]="scsi0:01:02:00"' 'ScsiDiskNames[2]="sdc"'

+++ ScsiDiskLocs[0]=scsi0:00:00:00
+++ ScsiDiskNames[0]=sda
+++ ScsiDiskLocs[1]=scsi0:00:01:00
+++ ScsiDiskNames[1]=sdb
+++ ScsiDiskLocs[2]=scsi0:01:02:00
+++ ScsiDiskNames[2]=sdc
++ local -i i=0
++ [[ 0 -lt 3 ]]
++ [[ scsi0:00:00:00 == \s\c\s\i\0\:\0\0\:\0\0\:\0\0 ]]
++ echo sda
++ return 0
+ diskName=sda
+ '[' -z sda ']'
+ local -i actualDiskSize
++ awk '$4 == "sda" { print $3 }' /proc/partitions
+ actualDiskSize=
+ '[' -z 0 ']'
+ ((  0 >= 121644148  ))
+ EchoToConsole 'ERROR: Disk '\''sda'\'' too small: 121644148KB needed, 0KB
available.'
+ '[' 'XERROR: Disk '\''sda'\'' too small: 121644148KB needed, 0KB
available.' = X-n ']'
+ _c=
+ _n=
+ '[' -t 1 ']'
+ echo 'ERROR: Disk '\''sda'\'' too small: 121644148KB needed, 0KB
available.'
+ tee -a /dev/console
ERROR: Disk 'sda' too small: 121644148KB needed, 0KB available.
+ unset _c _n
+ PerformAutoDDRSave

=================================================================================

It seems that there is a problem to obtain the actualDiskSize for sda, then i verify it with command 'awk '$4 == "sda" { print $3 }' /proc/partitions
' in shell mode. As a result, it shows the correct result. why this happens?

Can anyone give some suggestions ?

[gilbert08]

Are you restoring in different hardware?

[walker_yang]

No, restore in the same machine

[gilbert08]

If it is on the same machine it means that you will be using the same driver. Kindly recreate a boot cd after doing a prepare to restore.

[mandar_khanolka · Accepted solution]

>> 'ERROR: Disk '\''sda'\'' too small: 121644148KB needed, 0KB

This issue is due to problem in disk records mapping. I believe there is already engineering binary published by Symantec on this issue. Reference EEB ETRACK ID 2176010

I would recommend you to open a service request with Symantec to confirm the issue is same and get this EEB binary. Please make sure you provide below while opening service request with Symantec.

1. bmrrestore complete log (log containing above snippet)

2. bundle.dat for client. You can retrieve this from the client backup.

You can find this in /usr/openv/netbackup/baremetal/client/data/bundle.dat path in client backup.

Thanks.

Mandar

[walker_yang]

Thanks for your suggestion.

I saw this bug in document named "Symantec NetBackup™ 7.5 Emergency Engineering Binary Guide".

It's said that ETRACK ID 2176010 is a known issue that was identified, fixed, and available to customers in the form of EBB.

1. I verified BMR restore in the same enviroment using NBU 7.5, it revoled BMR restore failed problem.

2. On the other hand, i upgraded NBU using EBB of 2176010 in NBU master, and recreated SRT, then did BMR restoration, it went on well.

Hope this can help other somebody.