I have to come up w/ metrics to rationize the impact of adding clients to
our backup environment ie. one new client w/ 300 GB of data and X rate of
change will add X amount of resource load to a Solaris E4500 server. Trouble
is the resource load breaks down to the NIC, SCSI bus, memory etc...

I already have the canned Cricket time trending for CPU , memory, load etc..
and I have a custom Cricket app trending 10 DLT7000 drives and their thru
put.

I am now putting the bptm resource load in Cricket graphs. 

Does anybody have resources, etc... to help w/ this effort?
Surely my metrics crazed, purse string clutching managment is not the 1st to
ask for this. All this is to justify a HW refresh and expansion of our
backup environment.

thanks...