Enter the Aggregators

by Administrator on March 9, 2010

The other news item I alluded to in the previous post is a little company called Nasuni.  I chatted with them at the urging of the smartest PR person in the business, Judy Smith at JPR Communications, a couple of days ago and I thought I would share my views here.

Here is a pretty picture from the slide deck I was shown by the boss over at Nasuni.  Basically describes the play.  Nasuni creates a virtual NAS device under VMware that lets you connect to their partner cloud storage services in the background — including Amazon S3, Iron Mountain Digital, RackSpace and soon Nirvanix.  Essentially, these guys are service aggregators, so if you need more capacity, you can combine assets of different providers in a single virtualized storage pool.

I know, you wonder whether I’m off my meds for even using the terms “cloud” and “storage” together in a sentence.  I was wondering the same thing, actually.  Except this dovetails with something I have been wondering about for some time.

Let’s assume that a handful of the tens of storage cloud providers still exist this time next year.  Let’s assume further that they are able to compete with the falling cost and increasing capacity of disk drives (which will shortly see 10s of TBs on a single 2.5 inch platter with Bit Patterned Media and related technologies from Toshiba, et al.).  Remember that they are pitching $.25 per GB right now, but the Toshiba technology will reduce the cost per GB of disk to fractions of a penny if my projections are correct.

Anyway, one could imagine a world in which cloud based storage will have a fluctuating price, like any traded commodity. One day, service provider A would offer a TB at $10, the next at $9.87 or something.  His competitors would be doing the same thing.  Getting the picture yet?

Imagine that you buy disk at a spot price on a given day for one year.  You need more the next day, so you choose a different vendor with a different spot price.  Sort of what ENRON was trying to do with electrical power buys — before they got greedy and decided to park money offshore and lie to investors.

Anyway, you would need a common facility — a service aggregator — to get all the storage buys wrangled into a single service.  Nasuni does this in a very basic way, and adds a lot of bells and whistles to your storage in the process — the kind of things that hardware guys insist on putting on their controllers. 

There is nothing new about what Nasuni is up to — essentially they provide a virtualization layer and lots of functionality on that layer.  But, you got to hand it to Andres:  he is always thinking.

Now here’s where he and I stop our love fest.  Nasuni doesn’t aggregate or take any responsiblity whatsoever for the service levels provided by the vendors from whom his users elect to source their storage.  He doesn’t have the money to guarantee their uptime or compliance with SLAs.  Pity that.

Ultimately, cloud storage may sell to the masses of little companies who believe the woo the vendors are pitching and have no resources internally to field or manage a good infrastructure.  But, like ASPs and SSPs, I expect the appeal even to this crowd to wane as more and more companies can’t get at their data in a timely way when they need to. 

Interesting play, though.  Nasuni sort of reminds me of Xsigo in some ways.

{ 0 comments }

Everything We Need to Know About How to Screw Up IT…

by Administrator on March 9, 2010

We learned as kids.  Here is an article I wrote about it in Mainframe Executive, which has been cross-posted to MainframeZone.

I am also pleased that an article by moi — the REAL storage blogger wannabe — has just hit the wire at ESJ.com.  It covers file proliferation and the lame efforts of the industry to scale NAS to accomodate the data burgeon.

Frankly, I could have embellished it a bit further with two developments.  One I will treat here, the other I will save for another post.

The more I learn about it, the more I like the Novell File Management Suite.  NFMS is about policy-based file management based on user role:  a concept I like a lot.  I had a chat with them today and one of their really smart folks commented that in this DO MORE WITH LESS environment, people seem to be preoccupied with the WITH LESS component more than the DO MORE component.  As we have been ranting about at length here over the past couple of months, folks are still trying to throw hardware at the data burgeon — relying on thin provisioning, de-dupe, on-array tiering, etc. functionality joined at the hip to their array controllers to tackle the difficult problem of data management.  Technically speaking, data management has virtually nothing to do with hardware.

Oversubscribing storage via thin provisioning might make a dent in capacity allocation inefficiency, but it is downright idiotic to believe that it does anything whatsoever to economize on storage over the long haul.  Plus, as I have argued here, it exposes you to really embarrassing failures if an app ever does a margin call and asks for space it technically owns, but that has been allocated somewhere else.

We have already discussed on-array tiering, so I won’t repeat it here except to say that Tears of Storage do not substitute for real tiering, which should be based on a granular understanding of data and its business context, not mindless movements of anonymous bits across infrastructure.

Some proffer de-dupe as a substitute for data management.  I call these vendors the 420s, because I suspect they were out smoking a joint when the memo came out to do more with less.

The de-dupe pitch offends intelligence on many levels.  Once de-dupe is part of every file system, all of the software and appliances you are currently wasting money on can go bye-bye.  For now, it seems more sensible to do de-dupe in software, not hardware.  Here’s why:

Today, Sepaton announced a new class of de-dupe rigs.  They sell a software-only de-dupe package, but claim that some IT folks actually want a “one stop shop.”  So, they are doing the storage plus Sepaton “head” now, as well.  At $310,000 for a 30TB rig, that works out to about $10K per TB.  Since their software sells for $2500 per TB, that means that they sell a TB of commodity storage for $7500 — quite a bit more per TB than the $79 1-TB SATA drive I can buy on NewEgg.  In their defense, they say their cost is much less expensive than a roughly comparable Data Domain rig from EMC, which proffers a non-clustering (hence inferior in Sepaton’s view) box of rust for $15,665 per TB (not including 10GBe cards). 

I will let Sepaton and EMC duke it out over the numbers.  But I think the comparison is silly on its face.

Let’s look at the reality.  A spokesperson for a major telco spoke at a show I attended recently and observed that he was planning to ditch tape in favor of Data Domain VTLs, until the costs of management for the DD rigs bit him in the tuccus (you need to stand up another box of drives and manage it separately when you run out of space on the rust) and he discovered that he needed to classify data to determine which files to direct to the box so that the rig would be utilized efficiently (aka would give him the best de-dupe ratios).  He never dreamed that data management was a prerequisite for getting any kind of value out of that value add technology that hardware vendors insist on adding to their already bloated array controllers.      

Folks, the thing about data management is that it focuses on data management, not on capacity management.  Novell’s stuff is best of class in the products I have reviewed.  Dave Condrey’s team has done an out-fracking-standing job with this software, which lets you set policies for data movement based on user role.  Is it perfect?  Nope.  Will the Britney Spears files that that HR guy downloads at lunch be exposed to the same policies as the files created by his legitimate work effort?  Possibly, depending on how you write your exclusion policies.  But, now we have a sustainable way to tag a file, expose it to appropriate integrity, protection, and security services, and move it around infrastructure in an intelligent and compliant way.  AND IT IS ALL INVISIBLE TO THE USER.

The cost per TB?  No clue.  They sell software on a per user license basis:  MSRP is $37 before volume discount.  No magic mushroom inflation of commodity disk price there!

I would like to see any hardware vendor compete with this.  That is how you not only do the WITH LESS part, but also the DO MORE part.

IMHO.

{ 0 comments }

The Last Boyscout meets Mr. Milo

by Administrator on March 9, 2010

“What will it take to make you scream?” asks Milo, the bad guy in The Last Boyscout.

“Play some rap music,” responds Bruce Willis in the lead role of private investigator Joseph Hallenbeck.

A couple of posts back, I noted that StorageRap had taken it upon himself to decide who was a storage blogger wannabe.  I was chided in his blog, then he reiterated his pronouncement that Compellent’s blogger was a wannabe for “not engaging the rest of the storage blogosphere.”

Engaging with the rest of the blogosphere seems to be code for responding to Farley’s posts, I guess.  Seems pretty pointless to me, so I will happily join the world of the wannabes going forward.  So long, Mr. Farley.  Keep doing whatever it is that you do.  Rap makes me scream.

I completely disagree with most of the vendor subsidized bloggers out there who claim that on-array tiering provides value.  Farley and others claim that it reduces cost.  Maybe that is true in the short run, but I kind of doubt it.  Have you seen what the price mark-up is on those big capacity drives being used in tier 2?  And does the movement of data in a non-granular way to bigger, slower drives do anything at all to address the root cause of storage inefficiency that the pro-tierers like to cite:  the mismanagement of data itself?  I think not.

This is just a way to make the junk drawer bigger.  It does not solve the problem.

Yesterday, and this is a bit off topic, I chatted with some folks from Interlock Technology in Cambridge, MA.  I did the interview because I was intrigued by this company’s value prop.  They are using XAM — you know, that SNIA thing designed to create a way to place data in whatever SNIA certifieds as a XAM-compliant archive repository (modeled of course on EMC Centera) — not to place data on Centera, but to migrate it off the box.  I doubt that is what EMC had in mind when they participated in the development of XAM.  According to the CEO and CTO (the latter a former member of the Centera team who is now trying to buy back his soul before he dies), EMC boasts 5,500 Centera customers with about 600 PB of stored data.  EMC, said the CEO, did everyone the favor of raising awareness about content addressable storage (CAS).  True enough.  But they also delivered one of the worst products ever thunk up by man to host the data.  That’s my opinion, of course, since I don’t like proprietary rigs that turn storage into data roach motels.  There were so many other foibles in that product that I stopped tracking them, but customers were hamstrung because the only way to pull data out once you had committed it to Centera was to buy tens of thousands of dollars of EMC software.

That’s the problem that Interlock tackles.  They can extract 6TB of data per day from a Centera rig, which you can then migrate to Joe’s JBODs if you want.  Moreover, they work with the software you originally used to park your data on Centera, so you don’t lose connections to your data when you move it — or extra metadata you are storing with the data, such as retention/deletion markers.

I like what I’m hearing.  Interlock doesn’t reverse engineer the EMC API (something other vendors were able to do a while back, but were afraid to promote for fear of incurring the Wrath of Tucci), these guys are using XAM.  Hah!  I guess SNIA was good for something after all.

What really amazed me were the reports that Interlock’s software was able to provide on Centera.  They offer an assessment service where you get to see how the data is laid out in the Centeras you have and how all that retention, integrity and data loss prevention is working out for you.  They showed me one report from a customer who was growing Centera like crazy because the SE who installed it and set it up didn’t implement a retention/deletion policy.  The box was chock full of data that was supposed to be deleted years ago.  (Maybe that’s what causes the need to tier storage.  Whaddaya think, Mark?)

The CTO gave me some skinny on Centera issues, which I will follow up on later.  It was difficult to get a straight answer, however, to the question of what a Centera actually costs.  I have often been told that it is a drag-along, added to many a Sym or VMax purchase for less than list.  The units top out at a relatively small capacity for what you pay for, though EMC has added a Centera Virtual Archiver to the mix now that is essentially a redirector that lets you “federate” multiple heads.  The performance hit in a federated system is legend, perhaps accounting for why such a small number have sold.

There are about 200 to 300 Centera API compliant archive apps out there, I’m told.  Interlock is slowly but surely working with each one to ensure that apps will know how to read their data when it is migrated by their stuff from Centera to another platform.  We talked about the transformation process that your data must go through at some length, and I look forward to learning more.

I’m not in love with Interlock’s use of VMware as a hosting platform for their software, but I understand their reasoning.  You can buy their migration engine either as software preconfigured on a 1u rack server running VMware, or you can just buy the VMs themselves to install in your existing VMware setup.

Last note:  Gartner now says that VMware ain’t as shiny happy as everyone expected.  They place market penetration at 20%, which is well below what VMware expected after spending all that money on marketing and evangelizing over the past few years.  (IDC puts it lower, at around 15%.)  So much for that revolution…

{ 0 comments }

So Much for “Getting Along”

by Administrator on March 2, 2010

So, Sepaton has a funny on YouTube about a tape ejecting from a PC drive and killing Don, the IT guy.  I guess that this is an example of how a company uses “new media.”

Hmm.

To go along with this video, the company just released the results of a survey of 400 IT folks in the US and Europe.  The results support Sepaton’s Tape is Dead story (SEPATON is no tapes spelled backwards).  De-duplicated disk is the replacement, they argue — though privately they claim to work within the tape ecosystem.

The whole thing reminds me of something Adolph Hitler once said about propaganda:  blocked from entering the mind through the front door of reason, we will convince folks by going through the back door of entertainment.

I have a call scheduled with Sepaton on Thursday.

{ 0 comments }

Bling is Out, Tape is In

March 2, 2010

For tomorrow’s webcast at Redmond Magazine:

Hope some  of you can join us.  Link.

Read the full article →

Get a Clue, StorageRap, et al.

March 2, 2010

Unlike many of my fellow bloggers, it seems, I have been engaging with consumers — nearly a hundred in the last two weeks — and gathering their insights about this whole tempest in a teapot called on-array tiering.  That’s code for 1) establishing a tier of Flash SSD in a conventional disk array, 2) adding [...]

Read the full article →

Cheers for Digital Reef

March 2, 2010

It has been a little over six months since I did my first investigations of Digital Reef, purveyors of a great piece of software for sorting the junk drawer into class-based categories.  I interviewed them on camera for a multi-segment discussion of what they were about and how their technology worked for the first wave [...]

Read the full article →

Pardon Our Dust

March 1, 2010

While I was in Texas last week, first with a group of companies at dinner in Dallas with my friends at Zerowait, then as a speaker and participant in FujiFilm’s Global IT Executive Summit in Austin, I received emails from readers that my RSS feeds on this blog contained some crap that didn’t belong there.  [...]

Read the full article →

Trust

March 1, 2010

The blogosphere is chirping this AM about disclosures of affiliations between tech writers, analysts and pundits and the corporations about whose products they write. Turns out a few folks are ringers — on the dole of the very companies whose products they support. I am staying out of this one.
I happen to think that opinions [...]

Read the full article →

My Bad, ProStor

February 20, 2010

Curtis Chan called me Friday to nudge me to write something about ProStor, the guys who have made arrays out of removable laptop hard disks. They had sent me an external USB-connected unit and one hardened disk cartridge (about the size of a pack of cigarettes) just before Christmas and I promised to put it [...]

Read the full article →