Quibbles – Directive Management Redux

A while ago, I made a posting about a long-running annoyance I have with directive management in NetWorker.

This time I want to expand slightly upon it, thanks mainly to some recent discussions with customers that pointed out an obvious and annoying additional lack of flexibility in directive management.

That’s to do with the complete inability to apply directives – particularly the skip or the null directive, against “special” savesets. By “special” I’m referring to savesets that aren’t part of standard filesystem backups yet are still effectively just a bunch of files.

Such as say:

• SYSTEM FILES:
• SYSTEM STATE:
• ASR:

(And so on.)

In short, NetWorker provides no way of skipping these savesets while still using the “All” special saveset for Windows clients. You can’t do any of the following:

• Hand craft a server-side directive
• Hand craft a client-side directive
• Use the directive management option in the client GUI (winworkr) to create a directive to skip these styles of savesets.

OK, the last point is just slightly inaccurate. Yes, you can create the directive using this method – but:

• The created directive is not honoured, either when left as is, or by transferring to a more standard directive;
• The created directive is “lost” when you next load winworkr’s directive management option. Clearly it lets you create directives that aren’t valid and it subsequently won’t deal with.

Why does this suck? For a very important reason – in some situations you don’t want to have to back these up, or you can’t back them up. For instance, on certain OS levels and bitness using clusters, you will get an error if you try to backup the ASR: saveset.

This creates a requirement to either:

1. Accept that you’ll get an error every day in your backup report (completely unacceptable)
2. Switch from exclusionary backups to inclusionary backups (highly unpalatable and risky)

Clearly then the option is the second, not the first. This though is effectively removing an error by introducing poor backup systems management into the environment.

It would be nice if this problem “went away”.