10-15-2018 07:46 AM
Hi,
I have an appliance that has to be decommisioned but need to move the accelerated server backups to a new appliance. the problem is that the WAN link speed between the client server and appliance is extremely slow and would take 3 weeks to do a full backup and rescan a new accelerator log. I have an SLP on the policy in place to backup to existing appliance and then dupe the images to the new appliance. is there a way of migrating the existing accelerator track log on the client to use the new appliance without having to rebuild the track log?
Any help will be appreciated
Steve
10-15-2018 10:01 AM
The track log will register a new track if it finds the storage has changed. You can try and rename what ever is required in the folder stucture to trick it. Look at this note.
10-16-2018 12:19 AM
Riaan,
tried that already. didnt work. so despite all the existing backup images having been duped to new storage it still insisted on creating a new track log
10-16-2018 12:39 AM
Hey
I did never succeed in this endavour too... ;( not sure what I am doing wrong... In my case policy name was changed due to newer standard naming convention... And at the end I had to wait for new track being recreated...
Did anyone succeed on this??
10-16-2018 03:27 AM
Hello,
is a new appliance within the same NetBackup domain as old one, or in a different domain (so the image was AIRed)? I guess that redirecting accel backups to a different storage unit will work in the same domain only. Additionally, you should expire image copies from old one.
I can test this today, I will let you know then.
Regards
Michal
10-16-2018 03:50 AM
it is an existing appliance within the same domain.
10-16-2018 05:21 AM
Hello,
I have tested both the same and a different domain and it seems to be working. Fored Rescan is initiated, but for filesystem backups it should not be network intensive, it is only intensive between client and its storage. So I did:
- performed some accelerated backups to MSDP1
- duplicated these backups to MSDP2
- when the same domain I expired MSDP1 copies (maybe not necessary)
- in the same policy I created schedule pointed to MSDP2
- in accelerator track log path on the client I renamed <storage server> folder, in case of different domain also <master server> folder
- run backup against MSDP2
Now in the Job Detail a message
unable to validate change journal usage <reason=no previous backup>
orcing rescan, each file will be read in order to validate checksums
accelerator sent 14231040 bytes out of 2812057600 bytes to server, optimization 99.5%
So accelerator track info is reused and only very small portion of data is sent to network.
Regards
10-16-2018 05:24 AM - edited 10-16-2018 05:25 AM
Are you saying that copy or rename of the Storage Server name in the directory structure did not help? (That is after all all images were successfully duplicated to new Storage Server.)
/usr /openv /netbackup /track /MASTER /STORAGE1
Anything else that may be different? Like Policy name?
Can you see in bpbkar log that a new Accelerator track log is created?