> I have an Oracle backup split over 2 archive schedules due to the number of
> objects backed up.
>
> Schedule A starts at 20:00 and takes 40 Minutes.
> Schedule B starts at 20:10 and takes 2 hrs.
>
> Schedule B waits for Schedule A to complete.
>
> At this time this is the only client node backing up.

If both schedules are associated with the same node name, the ADSM central
scheduler itself will force them to run sequentially. I know of two ways to
circumvent this problem.

One approach is to register the same physical system under two node names,
start a client scheduler process for each node name, and associate one of the
schedules with each node name. Note that the client system will need two
dsm.sys files, at least one of which will have to specify a node name and
client port number explicitly.

The second approach is to set up a schedule with action=command and use it to
run a script on the client system. The script would spawn two background
processes to run the two parallel backups.

Once you get the two backups to start up concurrently, you will probably have
to confront the issue other respondants have mentioned: getting the two
backups to use different output tapes so that they can transfer data
concurrently.


The second approach