Is There Such a Thing as Too Many Netbackup Policies?
I wanted to ping the forum on this topic. The enterprise I work with we have a shade over one thousand clients that we backup. The enviroment, as it stands, has nearly one policy per client, so the policy count is right about 850. Personally, I prefer to group multiple 'like' policies together with a maximum of 5 to 10 clients. But I don't really see anything out there regarding a "best practice" where it concerns how many policies one should implement in their data protection environment. I looked through the forums and didn't see anything specific. The Netbackup Tuning guide also did not shed any light on my question.
So I come to you, Netbackup faithful, to see what your opinion is on this, or if you have a resource link that I could be directed to. If so, that would be much appreciated.
Thanks.
Comments
One policy per client
To us it seems a matter of preference. I was forwarded an article from Backup Central by a co-worker that offered new perspective of why an administrator should make one policy per client. The author (Curtis Preston) makes a pretty good case...
In favour of grouping
I don't think you can find a best practice about this topic. It's really what best for you. I personally standardize like crazy to be in control, limiting the number of policies as much as possible. I don't like the idea of having 2500 policies (different configured ???) in my GUI - But it doesn't mean other shouldn't do it.
You may also check corporate policies to see if there are demands that may you assist you taking the decision.
Assumption is the mother of all mess ups.
If this post solved you’re questions please send a gratitude by marking it as a solution.
It depends
it depends on the strategy...
I many times club the hosts if they are having common selection list, but if they are having different selections then you will end up adding exclude list for those. That times it is better to have different policy.
Also one more thing it is not best practice to mix DB backup with file level. Keep seperate policies for DB servers.
-- Deepak W (Kindly close the thread if your query is resolved)
The Reason for the Season...
Thanks all for your replies.
The reason we do things as we currently do is that management is very "huffy," if you will, about job failures. One thing that happens a lot is that servers go down for maintenance at various time. SInce netbackup doesn't provide a way to deactivate a single client within a multi-client policy for a specific time frame, we just put each client in one policy, and exclude the policy from running during that maintenance window. Hence, the method to the madness.
"I’m an early bird and a night owl. So I’m wise and I have worms."
- Michael Scott
You will have to tell management
That X number of failed backup tries isn't important - it's the time-frame without backup that counts. What's the point of counting backup failures, if you re-run that client during the day, and thereby secure the time-frame.
We only report failed time frames to our customers, not how many times we tried to re-run the job. Make sence, and happy customers ;-)
Assumption is the mother of all mess ups.
If this post solved you’re questions please send a gratitude by marking it as a solution.
Would you like to reply?
Login or Register to post your comment.