EMC + Data Domain = Now What?

Anyone that knows me or follows this blog understands I was never thrilled about EMC shelling out $2.4B for Data Domain. I didn’t like the deal at $1.8B. But what’s done is done. EMC kept the asset out of NetApp’s hands and it acquired a nice piece of property in Data Domain.

So what’s in store for EMC and its customers as a result of this acquisition? I want to cover three things here in this post:

  1. EMC’s data protection vision and what’s changed.
  2. What this deal means in the near term for EMC as a company.
  3. Some caveats for EMC and Data Domain customers.

First let’s do the vision thing. At EMC World in May, 2009, there was a session on De-dupe 2.0. In it, EMC put forth a really cogent vision about placing de-dupe (or data reduction) everywhere. Essentially, I inferred it was EMC’s intent to extract data reduction IP and place it throughout EMC’s portfolio, allowing all these technologies (source, target, single instancing, compression, file and sub-file level, etc.) to speak the same language so that when I move data around my infrastructure I don’t have to re-hydrate.

I thought this was interesting, powerful and unlikely to happen in the near term with so many disparate technologies in the EMC portolio (Avamar, Quantum, compression, etc.). Add to this Data Domain and the chances of this happening get even more slim, (except in narrow instances like Celerra). Too bad because the potential to do this and allow data reduction to be incremental and additive as you move data throughout the organization has some real appeal to customers.

I think the vision is still the correct one and perhaps with virtualization technologies it can happen over time but I believe this substantial integration effort will take a back seat to the Data Domain integration. In the meantime, I’ve seen some promising pitches from the likes of Falconstor, Permabit, Storwize and some others talking about bits and pieces of EMC’s vision. All powerful. All very long shots to unseat EMC as the market share leader at this point.

What’s Next for EMC and Data Domain?

Quick question. EMC just spent $2.4 B of it’s hard earned cash for Data Domain. Where do you think it will focus its energies and what will it stress to Wall Street?

Correct Sparky! EMC will micro-manage Data Domain and do whatever it takes to show how oh so smoothly the integration of Data Domain ‘went.’ I put went in quotes because mark my words, EMC will show Wall Street how it hit all its milestones and the acquisition went according to plan. Wall Street and the analyst community will parrot what EMC tells them and the world will just assume that EMC has done it again.

Now don’t get me wrong, EMC is pretty good at doing acquisitions. But this one won’t go that smoothly. The reason is that EMC will over pay Data Domain (and EMC) sales reps through 2009 to make sure deals continue. Frank Slootman is the man in charge (for now) and these two facts are a mandate for the Data Domain sales channel to be incredibly aggressive at the expense of EMC’s core data protection business.

Specifically, Data Domain will push its products (with EMC’s blessing) everwhere including into the heart of EMC’s core data protection business. To me, its fine to kill the Quantum business because that’s dead anyway. But Data Domain reps, along with its partners like Symantec will go hard after EMC’s Avamar data protection business and push Data Domain products in places where they may not be the best fit (such as in virtualized environments).

And while the world may never be aware of this because of the financial gymnastics that will take place to dress up the integration of Data Domain into EMC, customers should be extra careful about what they buy from EMC and where they put it.

Buyer Be De-dupe Aware

I’m heading out to VMworld next week and I will be extra sensitive to seeing how Data Domain and its partners (specifically Symantec) are positioning Data Domain products vis a vis Avamar solutions; and what the response will be from EMC-classic folks. Here’s why.

Data Domain and Avamar never competed directly in my opinion. Data Domain was by far the superior product for tape replacement because you just ‘set it and forget it.’ Very simple (from an IT process perspective) and extremely predictable.

Avamar on the other hand, required a whole lot of re-architecting of backup processes which didn’t make sense if all I was doing was trying to reduce my reliance on tape. Where Avamar always made sense to me was two places: 1) Remote/Branch offices and 2) VMware/virtualized environments. If the challenge is I/O throughput where backup windows are being squeezed because you can’t keep shoving more I/O through the network, then adding Data Domain disk targets to the backend won’t do squat.

Avamar software on the other hand goes deep into the file system to chunk data and more efficiently organize information so that it can be reduced before you move it around the network. Incumbent backup software like NetBackup from Symantec hasn’t historically done this and for sure target-based de-dupe systems are not the answer here. Now I understand the traditional backup software vendors are working toward this and that’s an interesting wildcard – but Avamar is way ahead of the curve on this one.

So for customers it’s dead simple. If your problem is you want to move to disk-based backup to improve recovery and reduce your reliance on tape; and you’re happy with your current backup software and you just want to plug a solution into your existing processes…Then call your Data Domain (or other target-based de-dupe vendor’s) sales rep.

If on the other hand, your challenge is VMware backups (e.g. keeping P->V ratios high) or remote office backups or even backup windows for large file systems, a source-based de-dupe solution such as Avamar’s is going to give you far better return on investment. Remember, that you’ll be doing some serious thinking about your backup architecture and you’ll not be simply plugging into your existing backup software paradigm of the past. You’ll need to have some deep conversations with Symantec and other backup software providers to see if they in fact can match Avamar’s capabilities in the near term.

The conversations you should be having with your vendor are around strategic fit. What’s the right technology to solve my business problem. Don’t let the hyperbole of an acquisiton and all the gymnastics associated with such moves cloud the fact that you have a job to do. Understanding the right tool for the right job is essential.

The rest is just noise.


, , , ,