Auditing Our NetBackup Environment

[greg_hall]

We are currently running NBU 6.5.3.1 as our enterprise backup solution.

I need to audit our environment to ensure we are backing up all of the required production data. My plan was to export the information (server name\backup path(s)) from the NBU master into a spreadsheet and then contact the system owners to ensure we are still backing up the correct data. If not, close the gaps.
But when we contacted Symantec regarding this, we were told it really can't be done. Well, not without a ton of scripting. Does anyone know how to do this, dare I say - easily? I can't imagine that this is not a common request.

Thank you in advance for your help, 
Greg

[J_H_Is_gone · Accepted solution]

See what you can do with the check_coverage script in goodies.

there are some issues with it... like if you have a cluster and you are using both physical server names and virutal server names but it can be quite helpful.

[greg_hall]

Thanks we'll give it a try!!

[Stumpr2]

Judy is correct but there are more things that you need to keep in mind.

NetBackup is mis-named. It should be called NetRestore as the backups are only as good the ability to safely and efficintly restore data :)

1. Databases should either be shutdown for imaging or else purchase, install, configure and TEST restore the NetBackup database agents

2. Timing is critical for a server that does batch jobs or data migration. In other words make sure that the data that is supposed to be able to restore is actually there when the backup image is created.

3. Remember to use the restoration time (time to recover data) as the determining factor for creating the backup image. Too many backup admins focus on making the backup window that they miss the whole point of the image is the restore. What is the best rstore procedure?

4, Retention level MUST be set by the person that is the data owner. Determination of using 1 month or 7 years retention is their responsibility of the person that "owns" the data. Our job is to ensure we meet their retention needs when we setup the backup policy

5. Partial backups are too many times wrongly acknowledged as successfull. Here is a clue. Partials are NOT acceptable until verified as such. You must determine why the backup was only partially successful. Determine the files that were not included in the backup image and discuss it with the data owner to determine what to do with failing files.

6. Failed backups need to be identified and fixed as soon as possible.

7. TEST, TEST, TEST the restores and prove that you can recover the data as required by the data owner.

 

[J_H_Is_gone]

#5 is the biggest point, and one that I spend a lot of time on.

I look at every SC 1 and ask the owners if that file can be excluded, if they are ok with only getting it every now and then, or if we need to work out a way to get it backed up ever night ( like shutting down a program or service).

Then I document their response
file name ok to exclude per so-and-so
file name ok for failure per so-and-so.

[Mau_mau]

We just want to know if there is an Audit log for Veritas NetBackup ver 6.5.3? This audit log should contain which user logged on to the application and what activities were done. We need this to collect all the server's audit logs into one repository server.

[sdo]

I have heard that acitivty auditing, logons, changes made, actions taken etc - and the audit tyrail/logging of such will be delivered with NetBackup v7.0.

Also, FYI the installation and configuration of NBAC (NetBackup Access Control) (formerly VxSS) will be very greatly simplified in v7.0 too.

[Trevor_Jackson1]

The admin console logs may help with your auditing needs. Try looking for logs in the /usr/openv/netbackup/logs/user_ops folder (or Windows equiv)