cancel
Showing results for 
Search instead for 
Did you mean: 

Catalog Backup fails after OS upgrade

Local_IT
Level 3

Hello,

Veritas Netbackup : V.8.2
Policy Type : NBU-Catalog
OS Master and Media Server : W2019 Std
Storage : Media Server (Policy storage) and local (Disaster Recovery)

 

This server (Master Netbackup) has been updated to W2019 (previously 2012R2) and since "Catalog Backup" not work anymore. All other backup (VM, physical server) working fine.

File are written in local folder (Disaster Recovery : Path) but deleted when job fail.

Communication (Netbackup Management -> Host Properties -> Media Servers) with Media Server are ok.
Local folder (on master server), who contains all previously catalogue backup, are available with full control for local administrators group and SYSTEM

A new policy (for backup catalog) have been created, but same errors.

 

Job :

Parent (48254) :
File list : "CATALOG_DRIVEN_BACKUP"
Status : "1: (2) none of the requested files were backed up"

Child 1 (48255) :
File list : "CATALOG_DRIVEN_BACKUP"
Status : "1: (2) none of the requested files were backed up "

Child 2 (48256) :
File list : "E:\Program Files\Veritas\NetBackupDB\staging\BMRDB.db"
Status : "1: (24) socket write failed "

 

Do you have, please, any idea to solve this problem?

Thanks.

 

 

6 REPLIES 6

Local_IT
Level 3

3 other try :

First : Create new "storage Unit" localy. With <Master> in "Media Server" field. With "BasicDisk" in "Disk Type" field. With <local path> in "Absolute pathname to directory" field.
-> Backup good (1 parent and 3 child).

Second : Create new "storage Unit". With <Media Server> in "Media Server" field. With "BasicDisk" in "Disk Type" field. With <Media Server local path> in "Absolute pathname to directory" field.
-> 2 child job good and 1 child job in error.


Job :
Parent (48275) :
File list : "CATALOG_DRIVEN_BACKUP"
Status : "1: (2) none of the requested files were backed up"

Child 1 (48276) :
File list : "CATALOG_DRIVEN_BACKUP"
Status : "1: (0) the requested operation was successfully completed"

Child 2 (48277) :
File list : "E:\Program Files\Veritas\NetBackupDB\staging\BMRDB.db"
Status : "1: (0) the requested operation was successfully completed"

Child 3 (48278) :
File list : "CATALOG_DRIVEN_BACKUP"
Status : "1: (24) socket write failed"

Spoiler
21 avr. 2022 13:35:39 - Info bptm (pid=2308) backup child process is pid 7096.6800
21 avr. 2022 13:35:39 - Info bptm (pid=7096) start
21 avr. 2022 13:36:51 - Critical bpbrm (pid=6192) from client xxxxbkp01: FTL - socket write failed
21 avr. 2022 13:36:51 - Error bptm (pid=7096) socket operation failed - 10054 (at ../child.c.1287)
21 avr. 2022 13:36:51 - Error bptm (pid=7096) unable to perform read from client socket, connection may have been broken
21 avr. 2022 13:36:53 - Error bpbrm (pid=6192) could not send server status message to client
21 avr. 2022 13:36:54 - Info bpbkar32 (pid=1248) done. status: 24: socket write failed
socket write failed  (24)

 

third : Create new "storage Unit". Select one <Media Server> in "Media Server" field. With "PureDisk" in "Disk Type". With <DiskPool-Media> in "Select disk pool" field.
-> All jobs in errors (like original configuration)

Nicolai
Moderator
Moderator
Partner    VIP   

Run command : vxlogview -p 51216 -X "jobid=12345"

Where jobid is the activity monitor job id.

Attach output text as a file to a post, do not bulk paste debug text, the antispam system on Vox will quarantine the message.

Has the windows firewall re-activated post OS upgrade ?

llang
Level 3

We had similar issues after Upgrading from 2012 R2 to 2019 with NB 8.2. The Catalog Backups sometimes failed. After a reboot, the NB Database was corrupted. Support restored the DB from the last known good state and we then had to install EEB 4008711 v1 since there seems to be a known issue with the Sybase SQL Version used in 8.2. This resolved the problem for us. In the meantime we've upgraded to 9.1.0.1 without any issue.

Local_IT
Level 3

Hello,

Thanks a lot for your answers and your leads.

I do not know how or why but this morning (FR time ^^) catalog backup didn't start on its own, and when I forced it manually backup was good (1 parent (48331) and 3 child (48332/48333/48334))


-Firewall on the server who has been upgrade (Master) was not been re-activate post upgrade.
-Attached the result of command "vxlogview -p 51216 -X "jobid=48331"", where "48331" is job ID of parent of job launch this morning (job ok)(job automatic with the usual policy).
-Attached the result of command "vxlogview -p 51216 -X "jobid=48254"", where "48254" is job ID of parent of job launch yesterday (first message of this post) (job ko)(job with new policy created yesterday for test).
-Attached the result of command "vxlogview -p 51216 -X "jobid=48218"", where "48218" is job ID of parent of job launch automatically yesterday (job ko)(job automatic with the usual policy).


I will force another backup this afternoon for see if it's still ok

Thanks again.

Nicolai
Moderator
Moderator
Partner    VIP   

Hi @Local_IT 

I took a look in the vxlogview text, but wasn't able reach a conclusion as of why the catalog backup are failing.

Have you considered opening a support case with Veritas ?

/Nicolai

Local_IT
Level 3

Thanks you to have looked at the logs.

I've relaunch Catalog backup and it is again ok. I'll see how it goes this weekend, and see monday if it was a false alarm ^^


No because last case I've open the 2021-12-16, get close 3 months later (2022-03-18) with "can't/want help you". And these 3 months have been filled with "Veritas Agent and get back to the client" or "Veritas Agent to contact the customer and discuss the problem" or "Veritas Agent to contact the customer and answer on..."

 

Thanks.