The complex choices of data archiving

Re-Define Archive

I have a difficult time recommending archiving products - I dont like archiving. Let's face the reason archiving even exists - we would rather keep everything "on-line" but we dont have the disk space. So, we look for products that help us move this data off our expensive disks, and on to inexpensive tape. That's the reason folks, running out of space - but we need the 1's and 0's, so store the stuff on tape. Arggggh !!

The problem though is not archiving, it is storage, archiving is a temporary remedy to a much larger issue. The underlying issues is traditional storage, that is - building RAID sets and then stacking them in a box. New GUI's and features dont fix the underlying architectural problem - expensive filing cabinets. This worked in an older time when 100GB's of data was a "boatload". That does not describe today's IT.

I have seen studies lately that show that up to 80% of purchased disk is wasted. Now, that space is of course "used" but in what ways? First, the traditional method of storing has us allocating each LUN (here, we carve out a "canister", substantially larger than our data set, to hold our data). For all the disk we purchase, about 20% of that disk actually is data.

Then, the next step - of that 20% of data, only 20% of it is actually "active" data. Think about it - only 5% of the space we have purchased is actually active !!!!

We need to build space to match IOPS requirements, and store everything else on SATA. The problem with traditional methodologies is that at the point of LUN creation, we need to determine what goes where. This doesn't work well and doesn't scale. As of today, only one vendor I can find is addressing this need - Compellent.

If I can keep my data on-line for decades, then I only need archive for specific needs, and it then becomes a valuable supplement to an overall Data Retention Strategy.

Address the storage, then, archive takes care of itself.

Paul Clifford
Davenport Group
www.davenportgroup.com