cancel
Showing results for 
Search instead for 
Did you mean: 

problem emm launch

G_D_
Level 3

Hi,

After an ungraceful shutdown of my backup server (electricity loss), i'm unable to restart netbackup.

First, i found my vxdbms.conf was corrupt :

11:01:53.639 [10028.10024] <2> read_legacy_touch_file: Found C:\Program Files\Veritas\NetBackupDB\data\vxdbms.conf; requested from(vxdbms_conf.cpp.1529).
11:01:53.639 [10028.10024] <16> ReadVxDBMS_ConfFile: Illegal configuration file:  C:\Program Files\Veritas\NetBackupDB\data\vxdbms.conf

It's wasn't a 0kb file, but with strange symbol within.

I recreate it with https://www.veritas.com/support/en_US/article.TECH48879

But, it's still impossible to start, emm, and i have this error when i try to ping the database :

13:25:36.281 [6964.7844] <4> ping_database: ODBC connection string=DRIVER=NB SQL Anywhere;CS=utf8;UID=dba;PWD=***************;LINKS=shmem,tcpip{PORT=13785};ServerName=XXXXXXXXX;DBN=NBDB
13:25:36.281 [6964.7844] <16> ping_database: ODBC connection failed.
13:25:36.281 [6964.7844] <16> ping_database: ErrMsg [Sybase][ODBC Driver][SQL Anywhere]Le mot de passe ou l'ID utilisateur est incorrect, ErrCode -1, Sqlstate 28000

I found it's possible to reinitialize it (http://www.veritas.com/support/en_US/article.000044417), but, i don't want thing to get worse.

Should i continue in this way ?

 

Thank for your help

G. DESSOLLE

 

OS : w2k12r2

NB : 7.7.1

1 ACCEPTED SOLUTION

Accepted Solutions

mph999
Level 6
Employee Accredited

Providing you run catalog backups, you should have a copy of the catalog NBDB files in /usr/openv/db/staging.

Copy the vxdbms.conf file back from there and see if that helps.

 

 

View solution in original post

11 REPLIES 11

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

Please log a Sev 1 call with Veritas Support.

mph999
Level 6
Employee Accredited

Providing you run catalog backups, you should have a copy of the catalog NBDB files in /usr/openv/db/staging.

Copy the vxdbms.conf file back from there and see if that helps.

 

 

mph999
Level 6
Employee Accredited

OK, windows path to staging dir may be a bit different .... 

NetbackupDB\staging I think from memory

nbutech
Level 6
Accredited Certified

Recovering from Staging or your catalog is the best way to get NBU environment up and runnnig

 

 

 

 

mph999
Level 6
Employee Accredited

If only vxdbms.conf is 'damaged' no need to revocer everything

G_D_
Level 3

Thank you for helping out.

With the version of vxdbms.conf from staging, everything restart (fine i hope...). I test some bakup and restore, everything is working.

Is there any software or utility to test everything is allright ?

 

mph999
Level 6
Employee Accredited

It is alright, if vxdbms.conf is messed up, NBU wouldn't be running ...

 

 

G_D_
Level 3

It is alright, if vxdbms.conf is messed up, NBU wouldn't be running ...

I realize it wink

This backup server is disk based, and i have doubt about my data files.

My doubts are based there ?

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
There is no tool that I am aware due to all of the components involved in a backup system. The best way to test/validate backup media is by running test restores. Run test restores from backups that happened on the day of the power failure. The OS would've run a filesystem check on all mount points at boot time. Any disk or filesystem errors would be logged to the OS system log. Modern arrays have battery backup to protect any updates in the cache at the time of power failure. Array logs should confirm. Backups that were running at the time of the failure should've failed with status 50 after all came online and NBU should remove incomplete images during next Image Cleanup. Any other manual checks for orphaned images will depend on type of backup disk - basic, MSDP, OST, etc....

mph999
Level 6
Employee Accredited

OK, why are you concerned about the data files - I presume by this you mean the actual 'images' stored on the 'disk' storage unit ?

What has happened that makes you think there could be an issue.

I agree with Marianne, there is no real way to check these files apart from restoring. 

 

 

 

G_D_
Level 3

The best way to test/validate backup media is by running test restores

I test some, and everything is ok.

 

The OS would've run a filesystem check on all mount points at boot time. Any disk or filesystem errors would be logged to the OS system log.

 Everything is ok.

 

OK, why are you concerned about the data files - I presume by this you mean the actual 'images' stored on the 'disk' storage unit ?

Yes. I have a puredisk diskpool. Because of deduplication, with little corrupt data, we can lose a lot backup.

 

What has happened that makes you think there could be an issue.

I have no log of the corruption of my vxdbms.conf. This is what makes me doubt the rest.

 

I rely on your experience, i will make other restore test.

Thank a lot for your help.

You can close the topic.