This week I had a chance to spend two days with a proactive and
forward thinking Storage team within a large organisation (I can't
provide the customer name unfortunately).
This customer had over 750 Terabyes of San storage across four EMC
Symmetrix Sans, covering production and BCP copies of all configured
storage.
A real eye opener was the level of duplication inherent in a modern
High Availability (HA) system. In this particular case 4 times the
storage was required – for every Terabyte of usable storage, 3
Terabytes are standing by ready for a disaster. The cost overheads to
maintain 75% more infrastructure must be huge for any company.
Leaving Systems administration five years ago, Sans were just becoming
affordable, NAS devices weren't really featuring and my organisation
was still relying on Raid 5 for most data types and Raid 1 for
databases (including Exchange). HBA's were so expensive that only
selected applications would have been candidates for a direct
connection to storage.
In that scenario, 33-50 percent overhead was required for redundancy
of business systems, granted that no secondary site was configured
with a DR copy of production data; how much is no downtime worth to an
organisation?
Any opinions are welcome, I want to learn more.
—
Sent from my mobile device
Leave a Reply