Jeff,

I do not have such big Netware servers - they've all been converted to NT.
What I can remember in terms of performance when you have quite big netware
volumes is to try two options - maybe together.
The first is incremental by date option - specified by -incr on the backup
command line or as -incr on the overriding client options on your schedule.
Then there is processorutil xxxx - this specifies the number of CPU slices
Adsm must use the CPU before releasing it for another process to use.
There is a final option where you can specify verbose no - that will not
write each file being backed up to the schedule log - this makes quite a big
difference.

The incremental by date option causes the backup to run a lot quicker
because there is less overhead than when running a normal incr backup. It
only looks at the date to decide to backup or not.

Don't blame the backup software here - I think your biggest problem is the
NDS on Netware - when backing up if anything in the NDS is not working like
it should you end up with the backup process waiting on the NDS to respond.

I'm not a netware guy - so maybe find out from your netware people what can
cause the backup process to perform badly from the netware point of view.

Sorry I can not give you more than this - but it should keep you busy for a
while.
As a last resort you can try putting on traceoptions to see what exactly the
client is spending it's time on.

Regards
Christo Heuer
ABSA Bank
Johannesburg
SOUTH AFRICA


> Help!
>
>
> TSM Server:  TSM 3.7.2 for NT
> Clients effected:Novell 4.X and 5.00 Servers using TSM 3.7.1 client
>
> Issue:
>
> I'm encountering major problems backup up these Novell servers with these
> clients   It's just 2 out of 20+ Novell servers, but they are highly
> visible.   What's happening is that it will sometimes take up to 52+ hours
> to complete an incremental backup of around only 2 gigs.   This is
> completely unacceptable.  We have other Novell servers that will do same
> amount of data in less than 2 hours.   It's somewhat sporatic, and does
not
> happen all the time, but when it does, the following appears on Novell
> console:
>
>
> ANS1802E
>
> Incremental backup of 'SYS': finished with XX failure
>
>
> <X.XXMB> [-  ]
>
> It will sit in this state for hours.
>
> At this point it is apparently attempting to access the "VOL1" volume (or
> "SDATA") volume etc.   Both of these servers have more than just the "SYS"
> volume..  These volumes are not that large, only about 60 gigs.  I have
> tried a variety of switches in DSM.OPT file including "memoryef y", as per
> Tivoli's suggestion.   Nothing seems to work.   I don't recall seeing this
> type of problem with the ADSM 3.1.XX client.  I've sent logs and more logs
> to Tivoli in attempt to resolve this problem w/o success.   Any insight
> appreciated.
>
> Thanks
> Jeff