> Hello everyone.  I've been doing a little research about
> the DRM module. It appears to me that we are already
> doing(although manually) what DRM would do automatically
> to gather information in preparation for disaster recovery.
>
> I understand that in a real disaster situation, there will be
> some manual procedure to be performed before DRM could
> actually take over. Assumption only.

DRM produces a recovery plan on disk on the host where the production ADSM
server runs. If you use DRM you will need to develop some sort of procedure
for copying the plan to removable media, vaulting the removable media, and
copying the plan from removable media to disk on the replacement host prior to
recovering the server. You will also need to make it possible to find the
latest version of the removable media after a disaster, either by labeling the
media or by printing some kind of index of recovery plan files.

We run our ADSM server under MVS. We copy the recovery plan to a second file
on the tape containing our offsite backup of the ADSM database. We print the
part of the recovery plan file that lists the volume serial numbers for
database backups. Some MVS sites transfer the recovery plan to a PC and write
it on a floppy disk. A lot of non-MVS platforms support floppy disk drives
directly, making floppy disks a relatively convenient way to transport the
recovery file.

A recovery plan is divided into stanzas. In an MVS environment, some stanzas
are human-readable instructions, some are batch jobs, and some are ADSM
macros. I gather that the structure is similar for other platforms, except
that the batch JCL is replaced by some counterpart appropriate to the platform
(for example, ksh on Unix platforms). If you use DRM you will need to develop
a procedure that reads the recovery plan and writes a separate file for each
stanza. We use a batch job that runs a Rexx program based on a sample supplied
with ADSM. I don't know what kind of sample code IBM supplies for other
platforms.

> I don't know much about the DRM module but I would like
> to know more about it.  It can be another option I can present
> to management for disaster recovery.
>
> Has anyone experienced or somewhat tested a real-life disaster
> in which DRM was used. If so, is DRM worth having.

We have used DRM in two disaster recovery tests. The tests were done under
conditions that provide a pretty good simulation of a real disaster. Our
vaulted tapes were shipped to a commercial hot site, some of our staff
traveled to the same hot site, and we recreated the ADSM server and selected
clients on hardware provided by the hot site vendor. DRM reduced the process
of recreating the ADSM server to little more than running a series of batch
jobs. We felt that we got our money's worth by not having to develop, test,
and maintain home-grown code for recreating the ADSM server.