cancel
Showing results for 
Search instead for 
Did you mean: 

Backup failing with exit status 42

edla_shravya
Level 4

Backups are  failing with exit status 42. I am attching bpbkar, bptm and  bpbrm logs. Could you please let us know what might be the issue here.

 

Thanks in advance

 

 

1 ACCEPTED SOLUTION

Accepted Solutions

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

I do not believe the issue is with socket special files only. 
The issue is with bpdbm timing out.

I would personally try everything that is suggested in the 2 TNs. 
Or disable TIR if there is not a very good reason for using it.
What do you have to lose?

My only other suggestion would be to log a Suuport call with Veritas.
They will need all logs at level 5.

View solution in original post

9 REPLIES 9

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

We first need to see the text in Details tab of the failed job in order to know where and what to look for in logs.

edla_shravya
Level 4

Hi,

 

Please find the log in the attcahment.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
Are full and incr backups failing? When last did a full backup succeed? We can see that this is an incr backup but because TIR file is old, a full backup is done: "Warning bpbrm (pid=7012) from client vnmhnned1-bkup: WRN - Old TIR file /usr/openv/netbackup/tir_info//NetBackup_file_info.vnmhnned1-bkup_FILES is missing. Backing up everything in /. " Have you tried without TIR? Is everything on NBU 7.6? Master, media server and client? All running Solaris? Which version? Client name indicates that you have a backup network in place. Are other clients on the backup network working fine? The problem seems to be with comms between bpbrm and bpdbm on the master/media server. bpbrm is waiting for bpdbm to acknowledge receipt of catalog info but is not getting it: 15:12:47.288 [7012] <2> db_end: no DONE from db_getreply(): file read failed 15:12:47.288 [7012] <2> db_FLISTsend: db_end_sts() failed: file read failed (13) 15:12:47.288 [7012] <2> db_flist_complete: db_FLISTsend() failed: file read failed (13) Do you have bpdbm log for the same period to see what happened here? Maybe overloaded master/media server?

edla_shravya
Level 4

And also on bpbkar logs I can see below warning.Could you please let me know whteher backup is failing due to below warning

Warning bpbrm (pid=7012) from client vnmhnned1-bkup: WRN - Old TIR file /usr/openv/netbackup/tir_info//NetBackup_file_info.vnmhnned1-bkup_FILES is missing. Backing up everything in /

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
Have you seen my post above?

edla_shravya
Level 4

Hi,

 

Yes this is an incremental backup.

If we disbale TIR option backups are fine.there is an issue with INCR backup , full backup seems to be working fine.

cat /usr/openv/netbackup/version
HARDWARE SOLARIS_X86
VERSION NetBackup 7.5.0.4
RELEASEDATE Sun Sep 16 11:43:30 CDT 2012
BUILDNUMBER 20120916

 

solaris 10

 Incr is failing as of now with erro code 42 and if I will run full backup ..it will be successful ..
and next INCR will again be successful , however again after few INCR runs , this will again fail with error 42.

Sorry we do no have bpdbm logs for same period. We have recent logs of another  policy. Please find the logs in the attchment,

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

 

 

07:42:51.244 [29879] <2> ImageReadFilesFile::check_tir_state: Need to wait to tir to finish
07:42:51.244 [29879] <16> ImageReadFilesFile::executeQuery: Exceeded retries waiting for image record completion
......
07:42:51.247 [29879] <2> process_request: request complete: exit status 249 the file list is incomplete; query type: 60

 

Although the backup does not fail wth status 249, this status code is covered in this TN: http://www.veritas.com/docs/000032194 

Please try the suggestion in above TN.

Do you really need TIR in this environment? 
This feature adds quite a bit of additional load to bpdbm on the master server.

I also found another TN for very old NBU versions, but may still be applicable and will not do any harm to create the DB_TIMEOUT and DBMto touch files :   (MAX_FILES_PER_ADD  too):
http://www.veritas.com/docs/000082098

edla_shravya
Level 4

Hi Marianne,

 

Thanks a lot for your response. I have gone through below TN provide by you.

http://www.veritas.com/docs/000032194 

But here I have one more query. 

In the TN its provided that as there are 28,662 socket special files  it  takes bpbkar about 2 minutes to pass the list of files to bpbrm over the single established socket but in the bpbkar logs what I have attached , I can see there are only 3 socket special files. Eventhough there are only 3 files why it will take long time to pass the files to bpbrm. Could you please let me know whether it will depend on any other parameters as well?

 

11/17/2015 15:11:11 - Info bpbrm (pid=7012) from client vnmhnned1-bkup: TRV - [/usr/openv/var/vnetd/bpcd.uds] is a socket special file. Skipping
11/17/2015 15:11:11 - Info bpbrm (pid=7012) from client vnmhnned1-bkup: TRV - [/usr/openv/var/vnetd/terminate_bpcd.uds] is a socket special file. Skipping
11/17/2015 15:11:11 - Info bpbrm (pid=7012) from client vnmhnned1-bkup: TRV - [/usr/openv/var/vnetd/terminate_vnetd.uds] is a socket special file. Skippin
g

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

I do not believe the issue is with socket special files only. 
The issue is with bpdbm timing out.

I would personally try everything that is suggested in the 2 TNs. 
Or disable TIR if there is not a very good reason for using it.
What do you have to lose?

My only other suggestion would be to log a Suuport call with Veritas.
They will need all logs at level 5.