Christophe's Corner

When EMC announced V-Max in the spring of 2009, it included statements of intent to deliver fully automated storage tiering, otherwise known as FAST.  Many years ago, when EMC was innovating in the storage business, the company rarely if ever pre-announced future products. Today, it’s common for EMC to sell a future vision because its current products lack innovation and are not keeping pace with the competition. V-Max is a perfect example of selling a future that isn’t ready for production, and without FAST, V-Max lacks any real innovation and gives customers little reason to consider it.

EMC is Late to the Party, Again

There are two important pieces of “FAST”:

1. The ability to move data between high performance expensive storage and slower inexpensive storage
2. The automation of data movement based on policies

EMC first announced that FAST v1 would be available Q4… and now? The latest update from Chuck is that FAST v1 won’t be available until ‘early 2010’.  See Devang’s (aka Storage Nerve) take on this new information here.

Once again, EMC is playing follower. We’ve had many exchanges about this around the theme of replication for example.  In this case, Hitachi introduced Tiered Storage Manager (TSM) to the market in 2005 (it was available in a previous incarnation under the name “Data Migrator”) and has a thousands of customers who have purchased and deployed it.

In 2009, Hitachi introduced Policy Based Management Service (PBM), a service offering to assist customers in implementing policy-based tiered storage automation.

Even when EMC does finally ship FAST, Hitachi’s implementation will still deliver far more value to customers. Why? Because Hitachi’s TSM provides a far wider range of storage options that include heterogeneous arrays from a variety of storage vendors, including EMC. This means customers can utilize existing storage and protect current investments.

Contrast that with EMC’s approach. Announce a totally new architecture requiring a forklift upgrade and promise to ship software to allow you to exploit the system sometime in the future.

Where’s the ROI?
The ROI of automated tiered storage management is that it eliminates the enormous staff overhead associated with moving data. What’s more, while storage admins can identify SLA shortcomings and manually tune storage arrays, they almost never will move data in an array that is over-performing, meaning often data resides on fast storage when it should be moved to less expensive SATA disk. The bottom line is automated tiering makes storage admins more productive and more efficiently utilizes disk space.

The combination of Hitachi technology (platform, TSM and PBM) provide vastly superior ROI when compared to EMC’s FAST and V-Max technology for the following reasons:

1. Hitachi’s offering is available TODAY well tested and proven in the field
2. It supports not only Hitachi storage but virtually every major non-HDS storage array in the market. As a result, Hitachi’s offerings provide the ability to mine value from existing and future assets
3. On a related note, remember that by using High Availability Manager provides the ability to non-disruptively move data between existing arrays and future generations of product.  My point here is that EMC effectively aligns automated tiering with capital investments in new arrays in the form of V-Max. At the same time, EMC’s migration is disruptive (i.e. DMX to V-Max) further limiting customer ROI.  See my previous blog about this.

When it comes to automated tiered storage customers have choices:

You can choose to acquire products that will solve your problems and work with installed assets, TODAY (Hitachi).

Or you can buy more hardware (e.g. V-Max) and wait until EMC ships FAST and hope that it works as promised….sometime in the future.  Personally, I’m in no hurry. And it sounds like EMC isn’t either.

In reality, this isn’t much of a choice, is it?