Keeping It Simple: Building a Storage Network

Enterprise Storage Forum content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.

As with all large projects, building a storage network is easier to do if you divide the process into smaller steps. But when you’re just getting started, it’s hard to know exactly what those steps should be. Marty Foltyn recently discussed with data storage network expert Bill Peldzus where to start and how best to proceed when constructing a storage network.

As with all large projects, building a storage network is easier to do if you divide the process into smaller steps. But when you’re just getting started, it’s hard to know exactly what those steps should be. I talked to a data storage network expert, Bill Peldzus, director of Storage Architecture at GlassHouse Technologies, about where to start and how best to proceed. I expected to hear the real scoop, since GlassHouse offers independent storage consulting services and does not sell or resell any hardware or software, and that’s exactly what I received from Bill.

Marty Foltyn (MF): Bill, let’s say I’ve just been given the assignment to evaluate my company’s current data storage setup and to recommend our next storage system. Let’s also say I’m thrilled to be trusted with the responsibility but am anxious about doing a good job. What would you recommend?

Bill Peldzus (BP): Well, first I would congratulate you and remind you that if your management trusts you with this business-critical assignment, you’ve already shown them you are up to the task. Then I’d recommend your new mantra: “Keep it simple.” Data storage today is complex, and it will be worse in the future, so your main job is to chart a path that protects data, keeps your company up and running, and enables the IT staff and managers to do their jobs.

MF: I’d definitely like to “keep it simple.” Where do I start?

BP: Step back a moment from just “jumping in” and make part of your plan be a process to build out your system in phases. You can be sure each phase is working well before you implement the next. You can also take advantage of developments in new technology.

MF: How do I know which phase to start with?

BP: Study your company’s goals and how data storage can support them. This will reveal problem areas that IT can address. Rank the problems and the dollar effect each has on your company’s bottom line. This will help you sort your most important changes from the changes that can wait.

MF: After I have my phases, how do I sort out the technology to use?

BP: Here’s where the industry can help if you continue your “keep it simple” focus. While the historic push has been on disk density, in 2003 storage infrastructure is becoming the target. The storage product landscape is expanding to include services as well as software and hardware. We’re learning about application integration and management, storage policy management, resource management, and system management. Then there’s virtualization, continuity, replication, and storage networking. All this is built upon the hardware fundamentals of arrays, libraries, disk drives, and tape drives. To sort through this, marry your vendors to your vision. Tell them your top three goals. Since you’ve already clearly identified your goals and your phases, this should be easy.

MF: Do I need to stick with one vendor, or will I be safe buying products and services from multiple vendors?

BP: Interoperability is still an issue (see PowerPoint slide). If we think of a bare-bones SAN, we have over 36 vendors to talk to, and these vendors combined can offer over 67,000 product combinations. Many of these combinations are supported, but they may not be certified. SNIA’s Supported Solutions Forum (SSF) has made a good start toward interoperability. SSF vendor members make sure their products work together and the vendors pledge to support their products as used in the supported solutions. But this is just a start. It is possible to install a heterogeneous, multivendor storage network, but to do it, you must enforce your “keep it simple” focus and make sure your vendors work toward your goals, which might not be theirs. You have a big communication job here.

Page 2: Building a Storage Network Continued

MF: Is a storage network possible in a phased program?

BP: In 2003 and beyond, the trend is “Spend to Save.” Analysts estimate that for every $1 of capital spent on storage infrastructure, $5 to $10 are spent on supporting that infrastructure. And yet, while you’ll be able to add applications and storage this year, you probably will not be able to hire more staff. Through policies, procedures, and process, a storage network will help bring in savings.

MF: Could you define policies, procedures, and process for me?

BP: IT operates between users and the technology to make sure the technology serves the needs of the users as they meet their business goals. IT uses policies, procedures, and process to perform its services. Policies state how IT will govern users. Service level agreements (SLA) state how IT will be accountable to users. Process defines how IT will work in a repeatable fashion across technologies. And procedures are training made site specific. Standardized, documented processes and procedures are key to successful SLAs. Emerging policy engines use storage management standards for better defined retention and archiving, as well as more manageable infrastructures.

MF: While I try to “keep it simple,” and since labor costs to manage storage are rising, would it be simpler and cheaper to run a storage network that includes just one type of storage — all disk, for example?

BP: If you compare the dollars it would cost to lose data to the dollars it costs to store the data, you’ll probably want to phase in different types of storage. Optimization of assets ensures the most expensive storage is reserved for the data that is the most important to the business success and survival, while less important data migrates to less expensive storage. This ties directly to your understanding of the business needs and the phases you established early in your IT project evaluation. Storage networks also enable savings. With shared tape drives, you don’t have to buy a tape drive with every server, and you can use your tape drives 12 hours per night instead of 4 or 5. And with a shared pool of spare storage, you don’t need to allocate spare storage to every server or application. Instead, you can distribute the spare storage where it’s needed.

MF: If I settle on a good backup and restore product, do I also need disaster recovery?

BP: The business driver, which you identified at the start of your program, provides a clear answer to this question. Each data protection method has a different purpose. Backup and restore provide data integrity; disaster recovery takes care of site failure. There are other data protection methods also, such as clustering, which provides high availability. Employ one or multiple data protection approaches according to your business needs tied to your phased approach.

MF: I know I’ll have to keep my old installation running side-by-side with my new one for a while before I cut over to the new setup. And I’ll have to do this for each phase. What are some other implementation steps I should plan for?

BP: Plan for how to get your data onto your new data storage while you run your side-by-side systems. Available methods include the simple backup tapes, the medium complexity mirrored copy, and the complex specialized data migration and virtualization products. Also remember you’ll need a troubleshooting phase. If you monitor for normal and train your staff, they’ll be able to do the initial troubleshooting. Also, remember to evaluate your maintenance contracts for consistency before you sign them. It doesn’t do much good to have a two hour response on a storage array if you have a four hour response on the fabric devices.

MF: You mentioned one of the benefits of a phased approach is taking advantage of new technology developments. That sounds attractive, but are there things to look out for with my “keep it simple” mantra?

BP: Using new technology is exciting because it can answer some of our pressing business needs. So you don’t get burned, carefully establish baseline requirements. Be sure you know what’s supported, and be very suspicious of outrageous claims. You’ll know them when you hear them. Also build into your request for quote process how to evaluate startups with great new technology as compared to solid, steady old-line technology vendors whose products aren’t as cutting edge.

MF: Thanks for talking with us, Bill. I think I see how the process requires study, both of my business and of the industry.

BP: My pleasure, Marty. This is an exciting time for IT.


See All Articles by Columnist
Marty Foltyn
of BitSprings Systems

Get the Free Newsletter!

Subscribe to Cloud Insider for top news, trends, and analysis.

Latest Articles

15 Software Defined Storage Best Practices

Software Defined Storage (SDS) enables the use of commodity storage hardware. Learn 15 best practices for SDS implementation.

What is Fibre Channel over Ethernet (FCoE)?

Fibre Channel Over Ethernet (FCoE) is the encapsulation and transmission of Fibre Channel (FC) frames over enhanced Ethernet networks, combining the advantages of Ethernet...

9 Types of Computer Memory Defined (With Use Cases)

Computer memory is a term for all of the types of data storage technology that a computer may use. Learn more about the X types of computer memory.