Restarting an NDMP Grp

Question

Hi,

I had some failures on our NAS box this morning, (backing up via NDMP), but when I restarted the job, expecting the failed directories only to back up, ALL the savesets are now queued for back up.

I restarted the grp via cmd line (savegrp -R ). NW ver is 7.4 build 187

Is this normal behaviour ?
Is it possible to only specify the failed directories ?

Tks,
Liam

amediratta · Answer

If you restart a group, it will not backup only the
save sets which got completed during the previous
backup job. If ny saveset was incomplete at the time
of group failure, Networker will backup the entire
save set again.

This is not entirely correct as it depends on restart window interval.

lguinane · Answer

Hi Anuj,that was my understanding, that restarting a savegrp would not back up previously backed up savesets.It would only back up savesets that not did not complete successfully.

ble1 · Answer

If you restart a group, it will not backup only the
save sets which got completed during the previous
backup job. If ny saveset was incomplete at the time
of group failure, Networker will backup the entire
save set again.

This is not entirely correct as it depends on restart window interval.

DavidHampson · Answer

Go to configuration; groups, select group, right click properties, advanced. Check the field Restart window which by default is 12 hours - it means if you restart more than this time later then it will treat the run as a completely new run of the group. You may want to set this to a different value - if you always want a restart to be a restart set it to 24:00

Jason20 · Answer

Check the field Restart window which by default is 12 hours - it means if you restart more than this time later then it will treat the run as a completely new run of the group. You may want to set this to a different value - if you always want a restart to be a restart set it to 24:00

I really wish there was some way to suppress this behaviour. It's simply not backwards compatible with the pre-7.3 notion of restarting a group.

In the environment here in most cases it's the worst possible default action. When someone helpfully restarts a failed group only to have several terabytes of pointless backup activity take place for the sake of a couple of failed savesets.

I'd rather it just fail and say why rather than do something I didn't ask for.

If only you could set the restart interval to 0 and disable it, or even to set it arbitrarily high would be fine, better than the manual hacking around required to perform a restart in the manner which used to 'just work'.

Jason.

Anandkrishna · Answer

I completely agree with David. And this holds true for all the backup groups not just NDMP.

NetWorker

Restarting an NDMP Grp

Was this post helpful?