The speed of your SAVF test will depend greatly on the type of storage you have on your server. I ran a test a few years ago to see if a virtual tape drive solution would perform well against my LTO3 library.
If I recall I got about 80 Mb/s on both devices. I had expected faster times for my virtual tape drive until I looked at my system’s hard drive performance. Once I reviewed my disk performance I realized that I wasn’t going to get better times without some upgrades.
I’d would recommend you and your boss to sit down and clearly define what it is you want to accomplish. If the goal is to implement a disk-disk backup solution you will need to look at your hardware. Do you have ample storage to make this work? Do you have fast drives? Do you have enough IO capacity to make this work? There are many third party products in the market place that may be cheaper alternatives than a huge hardware upgrade on you power system.
<a href=”http://publib.boulder.ibm.com/infocenter/iseries/v5r4/index.jsp?topic=/rzam4/rzam4virtualtapeusevt.htm”>Virtual Tape Drive</a>
My testing found that I could setup 2-5 virtual tape drives to run simultaneous backups and that would definitely outperform my tape drive… But the tradeoffs were big for any users expecting to get work done during the backup window.
My first observation is that it appears you are doing the SAVLIB to tape interactively, while you are doing a submit job to the QSPL subsystem. Would not a better test be if you either ran both sets of jobs interactively or by submitting them to the same subsystem ?
Just my 2 cents,
Hope that helps,