Cost of Doing Nothing
by David Merrill on February 27, 2009
I spent time last week in London and Ottawa with 3 very large clients (and one prospect) and then again this week with a large client in the Pacific Northwest. An interesting question came up several times (after reading my 33 types of storage costs paper) with these 5 clients – Why do we not include the cost of doing nothing? Is there not an option of taking no action? Does this imply these costs are free or of no impact?
This cost is also referred to as:
These arguments are being used relative to large government investments for stimulus or bailouts…. But I won’t go there today.
I did not include costs directly in the HDS Storage Economics methodology because I wanted to keep these money types that are measurable and actionable. There is always the option of doing nothing, but in that decision there will be costs (from the 33) that have to be applied. In the case of not upgrading an old storage array, you will (in the cost of doing nothing) be incurring many costs over time:
- Potential HW and SW maintenance
- Increased difficulty in tools, vendor support
- Microcode upgrade and support problems
- Increased part failure
- Power, space and cooling (which is higher in a per-TB basis than current systems)
Comments (2 )
rob finn on 13 Sep 2009 at 8:09 pm
Migration seems to be a small cost in relation to the others. I had heard it is closer to $10K per TB. Are you dividing that by 3-4 years?
David Merrill on 15 Sep 2009 at 8:26 pm
We use $7-10K per TB for end of life data (or array based) migration. These tend to be conservative estimates, but they include the dual maintenance overlap, power and space of both arrays, the software methods for data movement (tape) and the small army of people needed to do this effort.
Clients above 200TB of usable data are in a constant state of migration, and therefore this cost becomes significant. We see migration costs that approach 15% of the TCO.
Dave
