checkpoint restart process

[AlGon]

Hi everyone,

Netbackup 7.1.0.2 Windows 2003 Server

Can someone clarify this for me please.

How does Netbackup restart a job from a checkpoint if the job fails?  Does it rely on the the VSS snapshot in any way or it this completely separate from checkpoint restart process.

Many thanks

[Mark_Solutions · Accepted solution]

NetBackup works through a file system in a logical order when it backs it up (like going from A to Z)

When you use checkpoint re-start it makes a not of where it has got to at each checkpoint interval so that it has a record of that.

If it fails for any reason it can then kick back off from its last checkpoint

So if it would check point at D, H, L, Q etc, and if it then failed when it got S it would re-start from Q as that was its last checkpoint.

The last fragement of the image it had written from Q to S would be scrapped and re-written when it re-tried.

Very simplified overview but hopefully this gives you the basis of the idea behind it

[AlGon]

Yes, that's really helpful, thanks Mark.  Does it keep a record of where it's got too in a log somewhere?

I just wanted to make sure that if there are problems with the snapshot during the backup that it's not going to effect the checkpoint restart location.

[Yogesh9881]

it is always good to set checkpoint if you are backing up large amount of data with big backup window.

[Mark_Solutions]

It wont start the checkpoint until it starts wrting "files"

It is also not available for all types of backups (It cannot do SQL, Exchange etc.)

The System State / Shadow Copy Components are also not checkpointed as they must always be a full and valid backup - that is why an Incremental of them as always actually a full.

This describes a few useful things to note: http://www.symantec.com/docs/HOWTO34475

Not sure off the top of my head if the checkpoints go in the tryfile logs (netbackup\db\jobs\) or if it just uses the catalog itself