Client initiated Oracle RMAN backups and server side cloning, trouble?

I have a setup with NetWorker 7.6.3.5 running on Windows 2003. We are in the advanced stages of planning a hardware refresh of the server and storage node, as well as an upgrade to 8.1 SP1.

In the meantime, I'm reviewing the current setup, which I inherited some months ago. One of the issues I've found is that I can't find clones of Oracle RMAN backups that I thought were supposed to be clone to physical tape (from VTL).

All backups that are designated to go offsite go into our "SqlFile" pool. A lot of MS SQL backups go here, from the older NM SQL module, a handful using NMM and others are a simple file backup of SQL dumps. The commonality for the MS SQL backups is that they are all scheduled from NetWorker and they work reasonably well.

We have one single Linux server with Oracle and the Oracle NetWorker module which integrates with RMAN. The DBA handles this integration, I don't have first hand knowledge on how it's setup.

All I know is that Oracle RMAN kicks off and sends the backup to the same "SqlFile" pool as above for the MS SQL backups. The level of these Oracle RMAN backups are all "manual", obviously.

In NMC, a weekly cloning is configured to take all backups from the "SqlFile" pool with levels full or manual and send them to a tape clone pool. No filters for groups or clients, just the pool.

The process takes well over 30 hours to complete, using a single LTO3 drive. This configuration was in place when I took over, and I hadn't given it much thought until recently.

However, when I try to report older clones using "mminfo -avot -c oracleserver -q "savetime > 1 year ago, pool=clonepool" -r name,savetime,ssid,cloneid,volume,level

There are tons of clones missing that I expected to see in this report. There's a few in July and August, one in October and then just those I ran recently in January.

Is the clone process really so fragile that I'm missing a bunch of cloned save sets, or is there something I'm missing considering the Oracle server is

a) The only one running manual backups.

b) The client side configuration purges backups older than 5 days

c) Is my only Oracle and RMAN integration.

In NMC, I now have noticed that the cloning is flagged as "Failed" and I have a short list of failed savesets. The last clone had four save sets as "Failed", but when I run mminfo -q "saveset= " on these I get a report indicating that there are indeed two copies, one on the VTL and one on physical tape.

I'm just trying to figure out if there's something missing in my setup or if the cloning in NMC is broken in my version.

I have already written a script to use nsrclone to do the same, with some better logging than what I'm able to find using NMC. My deepest concern is that the clones I thought I had are infact non-existent. The retention time is set for 1 year, for business reasons.