When Data becomes (hopefully) Information

The storage insider enemy

If you were a storage manufacturer, and decided your market strategy would be developed through indirect channel partners, what wouldn't you want?.

I mean, this is not a letter to Santa Claus, imagine you're sending an email to the devil. Who would you like the devil to engage?. Yess, of course, not your loyal partners, but those who are playing multi-product quotes in customer's RFPs. They're just able to wipe out all those costly marketing and positioning efforts by just saying: "Oh, yes, Mr. Customer, you know... all these bits and pieces of junk are really all the same, but we can quote at lower prices on all of it since we buy sooo much...". Lovely.

Implicaciones del uso de los EFDs

Los denominados EFD's (Enterprise Flash Disks), han sido objeto de un vivo debate en el área de almacenamiento de la blogosfera. Desde que EMC anunció el soporte de esta tecnología en Symmetrix y en la gama CLARiiON, los competidores más directos, como NetApp, Hitachi e IBM, han adoptado posiciones que han variado con el tiempo, pero siempre críticas con la tecnología.

Posiblemente el hecho de que tanto NetApp, como Hitachi o IBM (incluso la obsolescente HP) hayan intentado sembrar dudas sobre las ventajas de los EFDs, tenga que ver con su incapacidad de ofrecerlos a los clientes por el momento. Lo cual no excluye que lo hagan en un futuro próximo, lo que redundará en la idea de que la crítica era interesada.

Acuñando términos despectivos como "Flashdance" o manifestando dudas sobre la utilidad práctica de los EFDs, los competidores de EMC han intentado ganar tiempo para poder poner sobre la mesa del cliente una oferta consistente que incluyese el soporte de estos discos en estado sólido.

Corporate SMBs

There are two main ways to approach the SMB market from the IT perspective. First one is being a provider devoted, born and brought up in the SMB market.

Probably this is a sensible approach for many service providers such as channel partners. But it's also been a common strategy for some equipment manufacturers.

Second way is suddenly finding there's a huge volume market out there, in the SMBs. That's, while being a high-end product manufacturer. SMBs are eager to perform well and, only sometimes, sensible to the improvements obtained by the creative use of technology. And the unavoidable investments required.

Working out the iSCSI issue

 

I need to work out some way to get around issues raised when discussing iSCSI implementations with customers. As I stated in the previous post, there's som kind of "lesser solution" feeling about iSCSI, as if the technology was only suitable for small retail shops. Which is untrue.

I made some findings, so I can list the main complains about the idea of iSCSI. Lets build up the list:

The Big Picture in Storage

Storage Technology is aggressively evolving, as data volumes handled by services, which have to be stored and backed up increase.

Don't get confused, it's not only a volume problem. It also happens that nature of information stored for any purpose is increasingly complex. So a whole new area of skills and expertise rises with overwhelming power in the industry. Someone would call those experts "informationists".

An informationist, as opposed to a traditional systems engineer, is someone capable of understanding the foundations of the nature of data, the organisational needs regarding that data and how it becomes information, offering value to the enterprise. With such understanding in mind, an informationist will be capable of designing a storage infrastructure, which will meet cutomer's business goals through the appropiate managament of the data accross its lifecycle. This lifecycle and all the features associated to managing, backing up, preserving continuity and offering confidenciality and integrity, are better known as ILM (Information Lifecycle Management).

Reviewing Risk Asessment

Let's face it, consultancy and asessment methods are sometimes based on the consultant's needs to justify customer's investments in their advice. Of course this statement does not apply exhaustively to all consultancy activities, but we must admit there might be a few cases.

Lets review the common Risk Assessment tasks executed to produce a Risk Assessment report, as a base for a Risk Management Plan and Business Continuity Procedures. The tasks involved include listing the associated threats, that is, the incidents that may cause losses in anyway, given a certain infrastructure.

So let's say the given infrastructure is the current customer situation, and the potential threats as I mentioned, are the unwanted incidents that may happen. Those threats, the experts say, need an associated value in the Risk assessment process. That is, assigning a probability value to each threat. The probability is how likely the unwanted even is considered to happen. I will explain roughly how this works, so everyone can understand my point of view. It won't take complex maths.

"Security Monkeys" (Torvalds says)

It seems to me a bit ungrateful to qualify security people as "Security Monkeys". Of course I believe some of them are bordering paranoia. For some years I even believed a paranoic could be a good profile for a CSO.

Torvalds seems now a bit upset with security people, specially the guys at OpenBSD. His words show, apart from being somehow fed up with the issue, a "tiny" bit of arrogance.

Lets remember then.

When Linus came up with his monolithic, technologically obsolete piece of software called "The Linux Kernel" (some strange kind of Holy Grial worshipped by unexperienced university students, and of course, that successful technology ideologist from the MIT called Stallman), what were the main arguments used to encourage the adoption of such solution instead of a modular microkernel-based, commercial product?.

Widget Flashing

I was talking yesterday to one of my customers. When I returned to my office, opened the laptop and had a look at the blogosphere which I usually tend to look at ( not very wide space to say, but specialized ), had a look at Chuck's blog once again. I was surprised to see I was almost about to write a similar post. He called it Feature Lust by Storage People. Well, my english is not naturally as good, so I have to find a workaround to get the big idea fixed in my post.

Widget Flashing is a literal translation of what I consider one of the most negative selling practices. You just have to wear a hat with the logo of a top IT manufacturer, and let those flashing concepts flow in the conversation, as natural as if you really knew what the problems of the customer are. After all, you don't need to get into deep findings, do you?, you represent something so big and clever, that you're able to guess, looking at your customer's eyes, what his problems (and solutions) are. Even more, you need not know the problems, you have the true solution for them; your hardware and licenses.

After one of those sessions, driven by the appropiate rep, you get the expected results.

Midsize to large Tape backups

I keep wondering why the industry insists on developing large and costly multi-drive tape libraries.

It's rather strange for me. Yeah, I know, you get a lot of $$ every time you renew your maintenance contract. That's for sure an attractive point to look at.

But, why do customers buy them?

They're expensive, they're complex, they need expensive software to handle them, specific drivers for SAN usage, licensed by drive, and worst of all... they use magnetic tapes!

Magnetic tape cartirdges are obsolete stuff, why then is still being used?. We are now hearing about new technologies such as storage based on SSD's and we still deploy magnetic tape backup devices. Are our customers gone mad?. I wouldn't say so.

Rationale

I've taken some time in reviewing the investments done by some customers in Storage Technology. Covering the last three to four years, the pattern shows common aspects, regardless of the market sector or the size of the organisation.

Most of the deployments made respond to a "fashional motivation", that is, as an aswer to the following quesion: "We must buy disk space, and everyone else is buying SAN widgets, why shouldn't we?". Other group focus on investment protection aspects ("if we keep on buying DAS, we will sooner or later have to deploy a SAN or a NAS, so, why not now?"). Others focus on the freedom to choose from (almost a commodity in the low-to-midrange) server suppliers, being independent of their propietary DAS parts and options. But I rarely detected a planned, strategic, value building approach to Corporate Information Management. In the other hand, what else could be expected?