The Storage Economist

We have seen a shift in the automotive world to hybrid vehicles with part electric and part gas-powered motors. There are points of efficiency to operate with batteries, and then again when horsepower is needed. The intelligence of the engine control determines when the car should operate on battery or pistons, thus removing the decision-making for the driver.

Some believe hybrids are a stepping-stone technology to the day when we can have all-electric cars. The time for the hybrid (if it is a transition technology) will allow for the electric re-charging infrastructure to be built out, for the price of electric cars to come down, for the battery to become lighter and cheaper, and for the technology to allow more miles between charges. If you have been in a Toyota Prius, you may be mesmerized by the graphics on the console that shows the battery, engine and drive train schematics.

The analogy of hybrid cars may clarify where we are as a storage industry with piston disks (mechanical SATA or SAS) and the new electric motors (solid state drives). We know that all SSD systems are powerful, consume less electricity and cooling, but are expensive and have a limiting factor with the amount of volumes available in the industry today.

Are in a hybrid storage infrastructure right now?

Recently, HDS announced Hitachi Dynamic Tiering (HDT) software, which performs the intelligence of moving data and can be analogous to the engine management function of a hybrid car. The storage engineer, DBA, or application does not want to decide when to run on SATA or SAS, and when to run on SSD. These HDT pools can include SSD, SAS and SATA disks in the same LUN, and the intelligence can move high I/O pages to SSD when needed, and keep the lower-access data in SATA when appropriate.

This diagram shows how 50% of the I/O can be reserved for Tier 0 SAS, and 80% of capacity of SATA used for 20% of the I/O. Tier 1 in this example is SAS, with 15% of capacity and 30% of the I/O.

The economics or cost of tiers is dramatic with this new hybrid storage tier approach. Let’s look at an example of a traditional 36TB disk configuration, with pricing, and other factors.

Now let’s take a look at a hybrid design, using a few SSDs for Tier 0, and then SAS and SATA.

I intentionally left out depreciation rates, labor, power, cooling and maintenance costs, as well as many other of the typical 33 types of costs that we use in Storage Economics, but the simple comparison is obvious:

You can do this same exercise with SATA, SAS and SSD in different tiers and have some manual data movement function to move volumes between different storage tiers. In this older approach, you only can assign a LUN to one disk pool type, limiting your flexibility. With HDT, you can have a LUN assigned to a pool of disks that include SSD, SATA and SAS. This is the engine control intelligence to allow the right data in the right place for the right cost.

I believe that capacity-based metrics and cost models are becoming secondary to the cost of transactions (reads, writes, IOPS, etc.), and this will present a new perspective on economically superior storage architectures.

We may not be ready to handle a full SSD storage infrastructure for most of our business applications, but this hybrid, intelligent approach brings a new level of economics to our tiered storage architectures.