cancel
Showing results for 
Search instead for 
Did you mean: 

DSSU SLP duplications queue following NBU7.5 upgrade

LostinSF
Level 3

Platform Win2K8-R2. Following upgrade of NBU7.1.0.3 to 7.5 all SLP duplications to both DSSU and tape fail to execute. Running nbstlutil stlilist -U -image_incomplete shows:

lifecycle <SLP NAME> is IN_PROCESS
Copy to <DSSU NAME> of type DUPLICATE is NOT_STARTED

Media servers were disabled when master was u/g to 7.5; then Media servers were upgraded. Also resulting from the u/g are massive speed drops for backups to both disk and tape.

Thoughts would be very much appreciated.

25 REPLIES 25

LostinSF
Level 3

"Something deeper"... That's what I'm thinking. Though I think also that I may be getting bitten on some of the less desirable configuration methods originally used in the 6.5 implementation. Time to get out the best practice documentation!

Thanks to everyone for their input. No panacea, but the personal knowledgebase increases...

pikachu
Level 6
Employee Certified

did you follow the upgrade guide? if so what does the SLP flag say.. still at 0? Phase 1 never finishes.... If so cat the bpdbm and search for "Anywhere"

 

IF that is it... ping me ... we can open a case to fix that OR wait until monday and check 75 known issues. the eeb for bpdbm will get posted!

 

MKT
Level 5
Employee Accredited

Until image conversion completes, SLP and DSSU Duplications are not allowed to run. 

Following the upgrade guide if phase 1 has not completed after twice than the estimated time outlined in the NetBackup 7.5 Upgrade Guide.  To check Phase 1 conversion status, please run the following commands.

On UNIX non-clustered master servers:
/usr/openv/netbackup/bin/admincmd/nbemmcmd -listsettings -brief -machinename masterservername

If the result shows SLP_DSSU_MIGRATION_STATE="1", phase 1 has completed.
If the result shows SLP_DSSU_MIGRATION_STATE="0", phase 1 has not completed.

On UNIX clustered master servers:
/usr/openv/netbackup/bin/admincmd/nbstlutil -active -backupid NetBackup_0000000001
If the result shows Cannot connect to nbstserv, phase 1 has not completed.

On Windows non-clustered master servers:
install_path\NetBackup\bin\admincmd\nbemmcmd -listsettings -brief -machinename masterservername

If the result shows SLP_DSSU_MIGRATION_STATE="1", phase 1 has completed.
If the result shows SLP_DSSU_MIGRATION_STATE="0", phase 1 has not completed.

On Windows clustered master servers:
install_path\NetBackup\bin\admincmd\nbstlutil -active -backupid NetBackup_0000000001
If the result shows Cannot connect to nbstserv, phase 1 has not completed.

 
Error

Bpdbm will show the following messages:

<32> process_old_dbm_images: Failed to import 'image' into the NBDB to be merged with the pre-7.5 Dbm_Image table AND mark_bid_failure failed
<32> process_old_dbm_images: Unable to merge DBM images from the pre-7.5 tables - rc 14
<32> unlock_image: image_lock mechanism failed
 
 
 
Cause

There are several defects that can cause this issue:


2696978: If there is an entry in DBM_Image but the flat file header is missing then the upgrade fails.

2698456: Upgrade can fail if images reference a client that has been removed from db/images.

Mark_Solutions
Level 6
Partner Accredited Certified

AH! - Finally hit the Late Breaking News!

http://www.symantec.com/docs/TECH182929

Genericus
Moderator
Moderator
   VIP   

I have this same issue.

Phase 1 of the upgrade has completed successfully.

The use of alternate read server is not currently possible with 7.5.0.3...

 

I have 18 fiber attached media server which write to a VTL. Backups run at their best speed without causing jams and drive contention. These are then duplicated to physical tape.

I have 2 fiber attached media server that read and write to that same VTL and all are part of a media sharing group. These 2 media servers are the ones that also are TAN to physical tape.

All my LAN based backups run through the media servers that also have physical tapes attached, and these SLP work fine.

The backups that run through my other media servers work, but the second step of writing to tape never starts, since the alternate read server is not the owner of the backup.

Changing the owner of the backup does not seem to resolve this ...

 

I am seeking an EEB or solution at this time, or notification that the alternate read server is no longer supported.

A review of the 7.5 release notes, 7.5 upgrade guide, SLP best practices and 7.5 install guide contain no mention of alternate read server not being supported. In fact the 7.5 unix admin guide includes instructions on how to implement it.

Ticket 418-897-688 has been opened with Symantec regarding this...

 

 

 

 

NetBackup 9.1.0.1 on Solaris 11, writing to Data Domain 9800 7.7.4.0
duplicating via SLP to LTO5 & LTO8 in SL8500 via ACSLS

Genericus
Moderator
Moderator
   VIP   

Applied EEB

# /usr/openv/netbackup/bin/support/eebinstaller.2858615.1.solaris

This program performs an installation for files related to
PET2719199, SET2858615, EEB1 on platform solaris.

 

replaced the nbstserv binary

 

duplications run!

 

now to see if they ALL do...

 

NetBackup 9.1.0.1 on Solaris 11, writing to Data Domain 9800 7.7.4.0
duplicating via SLP to LTO5 & LTO8 in SL8500 via ACSLS