cancel
Showing results for 
Search instead for 
Did you mean: 

Accelerator Backups failed with status code 14 and 84

Sisso
Level 5

Hello,

 

We are running Netbackup server 7.6.0.2 on windows server 2008 sp1 r2.

In our evironment we have some server in a remote location that everyday we backup those servers to an msdp disk on 1 media server.

In this backup we use the accelerator and client dedup funcions.

One of those servers stop to backup suddenly only his disk D and this is the error that we got from the Job detailed:

 A portion of data to be included from a previous backup (backupid = tserver_1406308898, offset = 588519424, length = 131072) has incorrect checksum (calculated checksum from backup image 693fd7af250d68385d3ef2e17f1baccb94610cdc, expected checksum e53dc59ba54b0902328a21a4a32f4a80df006064)
8/6/2014 3:52:37 PM - Critical bptm(pid=10384) image write failed: error -1: plugin error    

 

Thank you for help.

1 ACCEPTED SOLUTION

Accepted Solutions

Andrew21
Level 3
Employee

Hi, Sisso,

Following is the accelerator work flow:
o) After the 1st accelerator backup, the 1st backup image is stored on MSDP, and program generate a tracklog on client machine.
o) For the 2nd accelerator backup, it uses some change detection technology to decide a file is changed or not:
  i) if a file is un-changed, the 2nd image includes that file from the 1st image.
  i) if a file is changed, program compares the file with tracklog:
    ii) the file's changed part is transfered from client to server
    ii) the file's un-changed part is included from the 1st image.

For Accelerator forced rescan, it assumes all files are changed, and compares the files with tracklog. If a file (or part of a file) is consistent with tracklog, the un-changed part will be included from the 1st image.

Before any including, program checks the checksum consistence between client file and image on storage.

If forced rescan can not resolve the issue, it may caused by 2 reasons:
o) the 1st backup image is corrupted.
   Solution: 
      i) remove the previous tracklog on client machine (install_path\NetBackup\track\master_server\ storage_server\client\policy_name\backup_specification\track_journal.v2.dat).
      i) run accelerator backup to re-build tracklog and image.
o) program include wrong image.
   Solution:
      i) after re-build tracklog and image, the 2nd accelerator backup still fail. It must be some bug in accelerator.
      i) you need to contact Symantec for bug-fix.

Hope it help!

View solution in original post

8 REPLIES 8

BTLOMS
Level 5

what is the client OS ?

AAlmroth
Level 6
Partner Accredited

I have seen this issue in a few other threads here. Most seems to have a problem with the track log is being corrupt, or the actual image is corrupt to build the new full image.

https://www-secure.symantec.com/connect/forums/incorrect-checksum-ending-status-84

I think one of the reason why the full rescan option in the schedule settings is there. Try run a schedule which rescans the complete server, and see if that fixes the problem.

 

/A

 

Walker_Yang1
Level 5
Employee

hi, Sisso

Is your backup is accelerator backup for VMawre? if so, this is a known issue, but has workaround.

http://www.symantec.com/docs/TECH222588

If it's accelerator backup for file/folder backup, could you provide the following info?

1. which OS version and NBU version is client?

2. Your client is located at remote location? data is transferred throuhg WAN?

Thanks

Sisso
Level 5

The server OS is windows server 2003, and this is a physical server that his backup is over the WAN.

I already tried to run a backup with rescan accelerator but it didnt help and I get the same error.

A regular backup without accelerator is working fine, only the accelerator give me the error I published on this content.

 

Thanks

 

INT_RND
Level 6
Employee Accredited

Does the client software patch level match the media and master servers?

BTLOMS
Level 5

I have seen this with 2003 servers.

Walker_Yang1
Level 5
Employee
hi, BTLOMS This issue is caused by inconsistent checksum from backup image and track log. In this case, if you continue to try backup, it might still get same error. So if possible, move your track log(C:\Program Files\Veritas\NetBackup\track\master_name\storage server\client name\policy name\drive name\your specified directory) for this policy to another location, and try full accelerator backup again, this is equal to regular full backup, as it will send all backed up data to server this time. meanwhile it will create new track log, and in the followed backup, it has optimization rate. Thanks

Andrew21
Level 3
Employee

Hi, Sisso,

Following is the accelerator work flow:
o) After the 1st accelerator backup, the 1st backup image is stored on MSDP, and program generate a tracklog on client machine.
o) For the 2nd accelerator backup, it uses some change detection technology to decide a file is changed or not:
  i) if a file is un-changed, the 2nd image includes that file from the 1st image.
  i) if a file is changed, program compares the file with tracklog:
    ii) the file's changed part is transfered from client to server
    ii) the file's un-changed part is included from the 1st image.

For Accelerator forced rescan, it assumes all files are changed, and compares the files with tracklog. If a file (or part of a file) is consistent with tracklog, the un-changed part will be included from the 1st image.

Before any including, program checks the checksum consistence between client file and image on storage.

If forced rescan can not resolve the issue, it may caused by 2 reasons:
o) the 1st backup image is corrupted.
   Solution: 
      i) remove the previous tracklog on client machine (install_path\NetBackup\track\master_server\ storage_server\client\policy_name\backup_specification\track_journal.v2.dat).
      i) run accelerator backup to re-build tracklog and image.
o) program include wrong image.
   Solution:
      i) after re-build tracklog and image, the 2nd accelerator backup still fail. It must be some bug in accelerator.
      i) you need to contact Symantec for bug-fix.

Hope it help!