Green Storage Has Limited ROI, But Supports Overall Efficiency

I care deeply about the environment, certainly more than I care personally about money, so it pains me to say that in most cases, making storage decisions based on power expenditure alone is not rational behavior. The world is driven by economics, and the stark reality is that the cost of power is only a drop in the bucket compared to the amount organizations spend on acquiring and managing their enterprise storage systems. Maybe someday a consumption tax or cap and trade system will tip the balance towards more responsible consumption of non-renewable resources, but in the meantime, the pricing of power (especially in the US) doesn’t give much economic incentive for good behavior. In fact, according to a report Forrester published recently, the amount of money typically spent on electricity to power and cool a TB of storage is only about 1% of the cost of buying that TB of storage (or about 4% of the annualized cost of buying that storage given that you only have to buy the TB once every 3-5 years but you power it every year). So, unfortunately for the environment, power cost itself doesn’t provide a very strong incentive for storage efficiency.

Fortunately though, the things that enterprises can do to reduce their power consumption costs are often the exact same things they can do to reduce the capital and operating expenses of their overall storage environment. Focusing on improving utilization (measured as the quantity of data written divided by the quantity of storage on hand) and increased usage of dense drives are the most straightforward and effective ways to reduce hardware acquisition costs as well as power consumption. There are many ways to achieve these objectives such as thin provisioning, reporting and reclamation to improve utilization, and tiering or wide striping to enable more use of dense drives. Whatever the motivation — economic, altruistic, or a combination of both — organizations that put significant focus on their utilization and dense drive ratios are likely to spend less money and be greener at the same time. And that’s good for everybody.

By Andrew Reichman

Check out Andrew's research

Optimizing The Branch Part 2: Forrester's Take on Branch Office Consolidation

In the first IT Infrastructure & Operations video podcast, I discuss trends in branch office consolidation as presented at Forrester's recent IT Forum US event, with a lovely backdrop of the Las Vegas strip. Read the accompanying report here.

By Chris Silva

Check out Chris' research

IBM in talks to acquire Sun?

A big blue cloud overshadowed Sun’s announcement today unveiling their Open Cloud Computing platform. Media was a buzz today at rumors of a possible acquisition of Sun Microsystems by IBM. Still a rumor at this point, the story brings up many questions about how feasible this acquisition really is and if it makes sense from IBM’s perspective as well as Sun’s.

Here is Forrester’s take:

                        First and foremost this would be an acquisition based on market consolidation where a smaller vendor is acquired by a larger player, in an effort to grab a greater total share of a market where growing via margins is impossible. With Suns core business based in servers, a declining market, this seems a perfectly reasonable scenario to envision. The main value for IBM lies in the margin dollars found in the UNIX server market. IBM Global Services is one of Sun’s largest global resellers so it would be an opportunity for IBM to secure this very loyal customer base and increase its share of the UNIX market where all the margin dollars are found in the server market. Add in HP and Sun’s recent partnership around Solaris and you’ve got another big motivator for IBM to look closely at a potential acquisition of Sun to strengthen itself against the competition. Furthermore, IBM has wanted to get its hands on Java for some time as it is a key intellectual property for WebSphere and in terms of Identity & Access Management (IAM), IBM would also dominate the market from a revenue perspective with a Sun acquisition. Forrester believes a combined entity would compromise somewhere near 30% of a $2b+ IAM market. This isn’t the first time IBM has made attempts at acquiring Sun but so far all attempts have been unsuccessful. There also should be no antitrust concerns being speculated as IBM Rational alone has yearly revenue in range with Sun.

                        In terms of Sun, their growth strategy of appealing to developers through open source software — Solaris, MySQL, NetBeans, GlassFish — and hoping they will then turn to Sun for servers and storage won’t work in a global recession. The strategy depended on the rise of a new wave of Internet businesses — Web 2.0 — and expansion of Internet offerings of telcos and existing large service providers called “red shift companies”. In a global recession, neither camp is growing at the rate Sun needs it to. This makes a potential deal a possibility from Sun’s perspective.

                        Still, acquiring Sun would pose a number of integration issues for IBM Software Group. Firstly, IBM aims at executives while Sun aims at executives and beyond with a strong marketing investment towards developers. Secondly, both sides have fundamentally different business models. IBM SWG is not structurally set up to monetize support revenue stream model which they would acquire and very likely would have issues maintaining SLAs required to maintain it. Additionally, on the marketing front, a decentralized marketing model at IBM SWG would find challenges integrating a more centralized model targeted at grass roots like that of Sun. Although these challenges are not insurmountable they do offer a glimpse at several significant deterrents for IBM.

                        While IBM is the vendor which sparked the conversation, other vendors have made bids for Sun in the past and if IBM is interested it’s not out of the question to think others like Cisco and HP might counter. Regardless of the who, if Sun is acquired it should come as no surprise to see the winning bidder to shed a lot of the weaker products (and a significant number of redundant staff) in the process. From a software perspective we would expect Java, Solaris, MySQL, and Indentity Management Suite to live on with the new company with the possibility of GlassFish, NetBeans, Open ESB, and Java CAPS being spun off into open source projects if they are to continue on.
Given the current state of the market any acquisition rumor is food for thought and should be considered but this represents a deal that has just has many pros as it does cons and should be treated as such.

Thoughts on an IBM acquisition of Sun? Let us know what you think?

By James Staten

With Christian Kane, John Rymer, Jeffrey Hammond, and Stefan Reid

Check out James' research

Check out John's research

Check out Jeffrey's research

Check out Stefan's research

Knowing Isn't Hot But It Should Be

Last week Jason Newton at HP blogged about what his company thinks (or at least wants you to think) are the hot trends in the data center for 2009. He provides a good list that's less a reflection of what enterprise customers are necessarily doing but certainly what they should be thinking. Heck, his list reflects a lot of the tactics we discuss with customers every day in our inquiries and published research, such as in "Retrofitting Your Data Center for Better Capacity".

While you can certainly debate the priority of the items presented, there's very little here we'd disagree on (not sure DAS is HOT). At the center of all his advice though lies a subtle subtext captured best in his #3 - Knowing is HOT. Guessing is NOT. Newton writes, "Whether you’re talking capacity planning for your apps and virtual machines, the power and breaker size you need per rack or the storage for your data explosion, using the old ‘rules of thumb’ for quarterly budgets aren't going to cut it in 2009."

Absolutely right, Jason. But the issue is bigger than this. Enterprise IT Ops leaders are often flying blind on the overall cost of running the data center, what their average server utilization rates are, exactly how many servers they have, exactly how much storage they have and what percent of it can be used or shared. These baselines are necessary to truly know what level of positive effect you can have on your bottom line through consolidation, virtualization or any other efficiency improvement effort you plan to undertake. And this information is crucial if you need to deliver a business case to the CFO for investments in these efforts. And with all the consolidating and virtualizing going on today you'd logically conclude that, of course, knowing is hot. Everybody must be doing it.

Boy how we wish that were true. Unfortunately our data shows time and time again that most IT Ops leaders have a very limited understanding of their overall costs, let alone a complete asset inventory, utilization analysis or even an inkling of what portion of the power bill their consuming. Which means that most are making the case for their consolidation investments using the old "trust me" methodology. You know, "the old stuff is in efficient. The new stuff is far more efficient. So we can clearly save some money by moving to the new stuff."

Now I know that none of you are portraying your consolidation justifications this crudely, but without a baseline cost to work from, unfortunately a lot of the reasoning comes down to something like this. Vendors count on you taking this approach so you will use their ROI models, case studies and testimonials as your justifications. VMware likes to say that the average customers achieves 10:1 consolidations and has case studies showing millions of dollars saved. That's great. But will you really see those same results in your environment?

Don't take their word for it. Do the homework so you know exactly where your costs break down and which have the most acute inefficiencies. The best IT consultancies can conduct this analysis for you if you are slammed feeding the MOOSE. And a raft of good asset inventory and capacity analysis tools are out there that can not only help identify the baselines but serve as scorekeeper throughout the efficiency game.

Make Knowing HOT in your company so they savings you are chasing become a reality and IT becomes a real efficiency hero this recession, rather than a speculative one. 

Want help in making this happen? Come to our next IT Consolidation workshop, March 17th in San Diego, the day before our first Infrastructure & Operations Forum. 

By James Staten

Check out James' research

Egenera's Strategy Shift Is A Good Thing

Earlier this week, The Boston Globe reported that Egenera laid off short of 100 employees under the guise of the weakening economy, but there is more to this story. The reduction also reflects a shift in strategy to increase its focus on PAN Manager, its virtualization management software. Originally tied to its unique BladeFrame hardware products, PAN Manager was freed earlier this year and is currently distributed by Fujitsu-Siemens and Dell. As is often the case for hardware companies, Egenera's crown jewels are in this software and PAN Manager is one of the most mature, feature rich and enterprise tested of the virtualization software managers on the market. It has been hardening in investment banks, health care companies and other demanding environments for more than 5 years and supports Xen and VMware and runs on traditional rack or blade servers.

Since software is sold more effectively via the indirect channel (through distributors, OEMs, VARs and resellers) Egenera is able to downshift its more costly direct sales efforts by focusing on software sales going forward.

The company said it certainly isn't abandoning its hardware, as it will continue to fully support and enhance BladeFrame going forward.

The irony here is that the Egenera BladeFrame design is quite prescient. Modern blade systems used in best practice with networked storage, look and act a lot like the Egenera BladeFrame. And with the addition of advanced blade management software like HP Virtual Connect and Virtual Machine Manager and IBM Open Fabric Manager plus Virtualization Manager, these systems are moving towards the level of enterprise RAS capabilities that Egenera has provided for years. It might behoove a smaller blade system competitor to pick up the BladeFrame system from Egenera to better position itself against HP and IBM.

While certainly the economy has some connection to Egenera's cost cutting moves, the underlying strategic change is more telling of the company's leadership, new direction and focus on PAN Manager, which stands a good chance of being a hardware and hypervisor independent virtualization manager worthy of enterprise IT Ops' attention

By James Staten

Check out James' research

Can Force.com Help Consolidate Your Web Tier?

At Dreamforce today, here in San Francisco, Salesforce.com announced a significant, and seemingly long overdue, enhancement to its SaaS offering. They announced Facebook and Force.com for Amazon Web Services that are pre-integrations between their platform and these two other platforms. This new capability lets enterprise customers of their CRM solution (or any other AppExchange or Force.com) provide a public front-end to their instance of these services, directly from these services.  The big deal with these additions is that they let you tie third party applications directly into your Force.com applications. In the case of the AWS integration, if you have applications or services built in Java, the LAMP stack or native C code, you can integrate them with your Force.com apps.

There is no Salesforce.com log-in required and there is no separate web tier integration you have to host and manage. For example, if a manufacturer used Salesforce.com to maintain its price list and order entry, it could simply build a public storefront in VisualForce and this storefront would be served up directly from Force.com. Previously, if you wanted to leverage Salesforce.com data on your web site you had to code integrations between the Salesforce.com application in question and your web apps; and of course carry the expense of hosting, sizing, and optimizing this infrastructure for performance.

We certainly don't see enterprises migrating their entire web presence to Force.com because of this, but directly providing this capability from the cloud will let enterprises eliminate a portion of their web infrastructure and free their developers from having to maintain and troubleshoot the integration code that made this possible before.

In today's recessionary climate, anything that helps streamline IT operations and eliminates costs is a good thing.

There are caveats to this, of course. You have to build these web front-ends in Apex or VisualForce, Salesforce.com's proprietary language and UI builder, respectively, and clients will have to understand the security, process and workflow implications of this change. But if you already have skills in their tools today – which is probably the case if you are part of their installed base – then the transition should pay dividends.

By James Staten

Check out James' research

How Is The Economy Driving IT Consolidation?

There's been plenty of news on the state of the economy and how it's affecting everything from the length of the work week to the ability to travel, even Steven Colbert has been forced to make some changes.

It's been said, though, that it's always a recession in IT Operations. Companies look to the advantages technologies like mobile and remote access, unified communications, virtualization and IT consolidation as ways to reduce cost while keeping - and in many cases, expanding - functionality.

So what are your peers doing to keep IT budgets in check and save their organizations money during this financial tumult?  I'm looking to our research panel to find out. We're currently in the process of fielding a panel survey to assess the measures organizations are taking to reduce costs and make and I'd like you to participate. While this research is in the field, joining our research panel gives you the opportunity to share your best practices, a link to join is here:

Join Our Research Panel!

All of the responses will be anonymous and will be aggregated. You give a little and get a lot, lending some time and your insights into how your IT organization is run while gleaning cross-industry insights on how IT plans to weather the rough road ahead.

I'm looking forward to your responses!

By Chris Silva

Check out Chris' research

HP And IBM’s Blade Rivalry A Boon For IT Consolidation

HP and IBM are tossing barbs at each other in the blade server space this week with dueling management tools that greatly simplify administration, whichever platform you choose. On Monday, HP announced the latest iteration of its Virtual Connect technology and today, IBM finally unveiled its competitive offering, Open Fabric Manager (although IBM’s won’t start shipping until Dec. 21). Both tools let administrators pre-assign network and storage configuration settings that fail-over and migrate with the server and virtual server images running on those blades. They both also, in these latest iterations, let you manage these profiles across multiple blade chassis (up to 100 chassis).

While one can certainly quibble over which vendor’s solution is better, IBM argues more flexibility, supporting a wider variety of Ethernet and FC switch blades. HP claims Virtual Connect is more mature, feature-rich, and capable. There’s no denying this type of competition is good for enterprise buyers. Blades are more power and space efficient, provide greater configuration flexibility, and now are far simpler to manage than volumes of traditional rack-mount servers.

For IT consolidation purposes, blade chassis are arguably the best consolidation platform available today. Their benefits over racks of traditional x86 servers are obvious. But their fit is now extending past that of even UNIX scale-up systems. Large UNIX SMP servers have historically leveraged hardware partitioning as their means of slicing up a large server. Blades are essentially the equivalent of this capability. Yet there’s more configuration flexibility in today’s blade chassis, as you can mix any ratio of storage and compute, intermix FC, iSCSI, 1GbE, 10GbE, and InfiniBand, and even mix processor families. And there’s no migrating of software from one platform to another — a big win. Add in server virtualization software and these new network connectivity management tools, and you have a very compelling solution for consolidation.

While everything about each vendor’s blade chassis are proprietary, that shouldn’t prevent you from leveraging these solutions to meet consolidation needs. The blade market is a new and important battleground for server vendors and their latest innovations are highly likely to show up here first.

By James Staten

Check out James' research.