Static cost for cpu and ram

[eric_silberberg]

Is there a way to set static cost per cpu and gig ram? I can get derived costs using the allocation or utilization method by updating or modifying cost drivers. However our cost model spreadsheet has a few calculations that don't line up with the cost settings options. We are on VCPP for licensing, we are a public sector customer and sometimes have to charge self funded units a different rate than state funded units, we use a 10 year depreciation on some datacenter gear, etc etc. The vRops costing functionality is ok but not enough to replace VrB. Our disk tier rates can be static. The public cloud rates that we download are essentially static. Why can't our private cloud rates be the same?

I've looked at the excellent pricing article VM Pricing with vRealize Automation 8.1 and vRealize Operations 8.1 - VMware Cloud Management  by @bgordon but I don't want to maintain VrA just for a flat calculation.at this time.

[sxnxr]

We looked at the built in cost in vrops to give a static cost per cpu and memory for a VM. We have several different hardware models and all cost different so we wanted to average them out so we would get the same per cpu cost as an example no matter what the hardware was. we then wanted to use the over allocation ratio in the policy as a multiplayer so that an environment with a 1-1 cost more than a 5-1.

Unfortunately that is not how vrops works and was giving different per cpu costs based on the number of cores and there frequency in the hosts. EG a 8 node cluster with a total of 333GHz was more expensive than a cluster with a total of 450GHZ even tho the hardware costs were the same.

I ended up creating a dummy CDC called base rate and added three custom properties one for CPU, one for memory and one for storage. I took the average cost of a host and applied a 70/30 split between cpu and memory. added that value to the custom properties.

On my real CDCs i added another custom property for my multiplier

1	1 to 1
0.5	2 to 1
0.33	3 to 1
0.25	4 to 1
0.2	5 to 1
0.17	6 to 1
0.14	7 to 1

When we do out API call from out front end all we do is take the request VM CPU size x the base rate value for cpu x the multiplier

I also create custom groups for the VMs that use a relationship to the CDC and have a custom property added to every VM in the group with the multiplier. I then create super metrics to do the same as the API call to get the cost of all deployed VMs

Bit of work but does what we need. Good thing is we can change the 70/30 split and update the base rate CDC values to get whet we need the cost to be.