My Friday Rant: Storage As A Commodity

If you're a frequent reader, you know that I use a lot of my customer experiences as fodder for themes in this blog.  Usually, it's about a customer or partner aspiring to do things better.

Well, the world isn't entirely made up of aspirational individuals.  And, as a result, occasionally I get into customer situations where it's a more of a bare-knuckle agenda.

And, since it's Friday, I thought I'd share one of those stories.

The Customer Meeting

I flew out to do a few customer meetings.  Usually, there's one or two key discussions that justifies the trip, and then you fill up any empty slots with whatever pops up.  This was one of those "filler" meetings.  You set your expectations accordingly.

I found myself in the bowels of a non-descript industrial building in a dowdy conference room, a well-worn and coffee-stained office table in front of me.  The air had a peculiar smell.

Across the table, three people: the head of procurement, what looked to be his apprentice, and someone introduced as the "IT manager".

I looked over at the rep with one of those "what the heck have you gotten me into?" looks that I save for these special occasions.

Turns out there was some sort of a storage transaction in play -- I didn't know the details -- and I guess I was there to attempt to defend EMC's proposal.  Now, I can usually do that quite well, given a bit of time and context, but that wasn't an option.

Within 3 minutes, the gauntlet was thrown down.  "Storage is a commodity" the head of procurement asserted, "and why should we pay any premium whatsover for what you're proposing?".

I paused for a moment, and said "that's a very fair question".

How You Think About Storage Should Change

I started by setting a baseline -- in many ways, storage in an IT setting hadn't changed much in the last 10 years.  We all used basically the same physical disk drives (FC and SATA), we all had either dual-controller or multi-controller designs, we all did local and remote replication, we all had some sort of management console to run our stuff, and so on.

Sure, there was interesting differentiation as you got closer to the various vendors, but -- from a safe distance -- they all looked pretty much the same to the uninterested observer.

But, I said, every so often the technology fundamentally changes.  This particular transaction represented the new regime vs. the old.  

And, as a result, perhaps your thinking should change as well.

Storage Media Has Changed

No surprise, enterprise flash drives are a new ingredient.  Sure, they're faster -- but they also can be cheaper as well -- especially if you've got applications that need a dollop of performance.

I asked them if they ran any databases (of course, they did) and I said that we'd be running those databases on a more cost-effective mix of a small amount of flash and the majority on far cheaper SATA drives.  

That was going to save money on equipment as well as power consumption.

As I explained what FAST was, and how it worked, I started to notice some uncomfortable looks from the IT manager.  

Turns out he wasn't entirely comfortable with this whole concept, so he had insisted on an old-school 15K FC drive configuration.  No flash.  No SATA.  No auto-tiering.

Storage Connectivity Has Changed

Rather than get stuck, I kept going.  

How you connect servers to storage has changed, I said.  Rather than use traditional (and expensive) FC everywhere, our unified storage platforms now support a flexible mix of FC and ethernet protocols: NFS, CIFS and iSCSI.

 Use FC where you feel you need to, use ethernet everywhere else -- and save money and effort in the process.

By now, I was watching the IT manager for another uncomfortable look.  Yep, there it was.

Yes, EMC had suggested that, he said -- but he felt more comfortable with FC SANs, and wanted to use separate storage devices for his file servers.  I said that was understandable, but hardly necessary these days.

And that was an expensive decision -- more cost, more effort, less ability to pool and share, etc.

By now, there are some glances going back and forth between the head of purchasing and the IT manager.  The apprentice type was staying out of it.  The rep had started to relax a bit.

Time to keep going ..

Storage Management Has Changed

I was starting to get on a roll by now.

The way you want to manage storage had changed as well, I asserted.  Not only were there much better new management consoles like Unisphere, but we were moving more and more of the storage stuff into the VMware environment, where the server admins could do most of what they needed without getting a storage person involved.

Less effort, faster results, fewer mistakes.  

By now, the IT manager was getting a bit agitated, and started to respond.  He said that his team had looked at Unisphere, but it didn't work with all the home-grown stuff and processes they'd developed over the years.  Fair enough, I countered.

He also said that he thought it a bad idea to have VMware admins provisioning and monitoring storage by themselves without any supervision from the storage team.  Chaos could result.

I said, yeah, but of course there'd have to be a bit of new process and procedure involved, but it was now generally considered a best practice to have the storage people set up the pool of resources, and simply turn it over to the VMware administrators so they could get their job done.

The reason people did it was there was far less effort and complexity involved, freeing up the IT staff to go work on more important stuff.

I definitely saw what looked like a stink eye come my way from the IT manager.  Sorry, dude.

Stepping Back A Bit

I finished off my tour by saying there was a lot more to the story along the same lines.

How we do backup and recovery had fundamentally changed, thanks to data deduplication.  How we do business continuity and remote replication had changed as well.  It was quite a list to consider.

I said that we as vendors were spending enormous sums to make storage faster, more efficient, easier to manage and do a better job of protecting your data.  The stuff was already in the market, and had proven itself over and over again.  Things had changed for the better.

Strip all that newer stuff away, and -- yeah -- parts is parts.  No arguing that.

I gently suggested that maybe it would make sense to do a quick professional services engagement to make some recommendations as to how the IT team might better take advantage of what's now on the market.  

Use EMC products, use someone else's -- maybe it was time for a process upgrade.  

General agreement all around.

And I thought to myself -- traditional thinking, yeah, now that's a commodity :-)

Comments

[Greg Knieriemen]

Another great post Chuck. One of the philosophies floating around that leads to comments like "storage is a commodity" is that the abstraction layer that VMware creates will ultimately manage the entire stack (servers, switches and storage) and render each of those parts as commodities. Now EMC is way ahead of this with Ionix Unified Infrastructure Manager 2.0 but don't management tools that virtualize the stack (not specifically Ionix) have the potential to yield those parts as commodities?

[Chuck Hollis]

Hi Greg

You're quite right: today's differentiation is tomorrow's commodity. That's been the case in IT for at least the last 20 years, and perhaps the next 20 as well!

That being said, there always seems to be room for more differentiation, more integration and more value-creation in the stack or solution as different parts commoditize.

Smart vendors (and smart customers) will recognize this trend, and hopefully plan accordingly!

-- Chuck

[StoragePro]

Mike from NetApp here

I enjoyed this very much. Lots of us have been in the storage industry for a long time - 15 years in my case, (good Lord!) and we have seen change happen at an amazing clip. In fact, 15 years in enterprise storage would put you in the industry from just after its birth! If I go all the way back to Middle school - 7th grade to be exact, my first programming efforts were stored on paper punch tape! 5 years later, we had Apple II's and a floppy disk holding 16K. Ah, those were the days! Today, I store a lot of my stuff in the "Cloud". Nice.

Today, the acceleration of change is bewildering! I think some of the reticence to change comes from good old human nature (fear), and the rest comes from a sneaking suspicion that they are being sold rainbows (more fear). Not long ago, as an SE at EMC I could sell a TB of disk (just disk!) for a million dollars. Now, you can buy 2 TB in a single drive from best buy for a few hundred bucks.

During my new hire at EMC I vividly recall Jeff Goldberg addressing us. He spoke about "Institutional Thinking" as EMC's biggest competitor. I still have my notes from that talk he gave, and it was riveting. Few people like Jeff ever walk this earth. The truth is, that fact has not changed one bit. It is not only our biggest competition, but it is our customer's biggest threat!

The one constant in our lives is that the biggest barrier to useful change is institutional thinking, be it for personal growth, or change that propels a business forward. Your discussion was a great example on how anyone - especially an IT manager, can overcome and break down these barriers to progress.

Every IT shop I've ever visited - 1,000's of them - have some of this friction holding them back from extracting the full value of the pieces and parts they've selected to use. I look forward to the day when the majority of business look at their IT infrastructure and staff, and smile, rather than grit their teeth.

More than anything else, the Human Element catches my attention as the driver or inhibitor to growth and progress. The tools chosen can make a difference, but more important to me is what you do with what you have that that makes the difference.

Cheers. My best to you and your family.

[Steve Ennis]

I'm glad I found your blog Chuck. This is fantastic insight into the marketing/strategy of EMC's products. I have an OK understanding of how some of the older products work, but I'm fairly new to the company, and do not know much about what OTHER companies offer with our new product line. I'm currently in grad school at Northeastern; They want us to apply some of the applications we talk about to our own companies business plan. I hope to dig deeper into the decisions EMC has made in the past, and how they plan on moving forward with an expanding market.

[Duane A. Burman]

Chuck,
I enjoyed your "rant"; as it was just the type of front-line interaction/confontation with the "handson" decision makers that so many upper management players like to avoid---reason I always tried to stage some "dry runs" to see if I really wanted to risk taking my own Sr. Management out to see a Client! I was Co-Founder of the Twin City CIO Council @ Y2K and had the "good fortune" to get to pull together Fortune 1000 CEO's, CFO's & CIO's with widely varying levels of understanding of their existing technologies, available technologies as well as even their own business operations! (one CEO told me aside that "He never allowed his CIO @ their Strategy Table; as he bored them with bits & bytes they never understood & showed no concern for learning where the CEO, SVP of Sales, COO & CFO were coming from---hence he made the CIO report to his CFO; so he could control him with a tight hand on the I.T. budgets!")
Today's CIO's are being rapidly sorted out by the need to "Run I.T. as Business"; and I truly hope that I.T. Mgr. recognized the great management mentoring you provided, Sir!