iSCSI vs. Fibre Channel - Not a Winner-Take-All Battle

These days everyone seems to want to talk about iSCSI as an up-and-coming challenger to the dominance of Fibre Channel SANs. In my view that's a myoptic view of the issue. This isn't an epic winner-take-all battle.

The availability of iSCSI SAN solutions with robust functionality and scalable performance certainly changes the landscape, but perhaps not how, or as much as everyone would like to think.

In the "old" days (a relatively small amount of time in our industry), Fibre Channel SANs were the only alternative for storage. However, the cost and complexity of the solution imposed natural limitations in terms of the "affordability" of these technologies. It found its natural place in the ecosystem of business solutions.

iSCSI SANs crawled from the technical primordial ooze somewhere along the line to join the food chain, and were immediately labeled as a low-end storage alternative not suitable for anything but limited use. Granted, in those early days iSCSI did have a lot to prove - and needed to "evolve" more to deliver all the functionality required to be taken seriously - and to prove itself worthy in terms of performance and scalability. Fast forward to today where iSCSI SANs have crawled a long way up the evolutionary ladder in a short period of time and are now encroaching on the previously unchallenged domain of Fibre Channel. Everyone wants to talk about the battle that surely will result.

Does anyone else think this sounds like an episode of Animal Planet?

The fact is that both iSCSI and Fibre Channel have both been evolving. The Fibre Channel industry has done a lot to address the complexity issue - and naturally prices have come down over time. While iSCSI was crawling up the food chain - Fibre Channel has been busy crawling down it as it offers its value proposition to new segments of the market. So we have two "species" moving towards each other and picking up speed. Now that's an episode of Animal Planet everyone wants to watch!

The animal kingdom/evolution analogy isn't a bad one to look towards for guidance in this situation. Two similar species, each with strengths and weaknesses, competing in a common environment. What's the result?

Well, in nature, species compete based on their strengths and a natural balance is struck. In the case of iSCSI vs. Fibre Channel, we already know that iSCSI will dominate the lower-end of the solution scale (price, low complexity, common technologies, etc) and Fibre Channel will have the upper-hand in the high-end storage solutions market where it's complexity and cost will be justified. But what about the middle ground?

That's where things get interesting. In this instance, we're going to have a rather large middle-ground where both species will compete within the ecosystem - in many cases head-to-head. Call it the mid-market or whatever you'd like. It's an area where the competition will be intense because the customer will have two viable technology alternatives for the same business requirements. It will be a changing landscape - with the fortunes of the two competing "species" changing as each continues to evolve - iSCSI fighting its way further up the food chain, and Fibre Channel fighting to move down the chain.

The good news - well - the customer wins. Any time the customer has multiple technology alternatives to solve the same business problems the customer comes out ahead. We vendors will continue to work hard to improve our 'species' to compete better for the available food supply and the customer will sit back and reap the benefits. Let's face it, the customer is the top of our food chain - and nothing is going to change that law of nature.

Hone your SAN skills online

I have to confess that I had never heard of the Regis University Online before, which proves one more time that I don't know everything and don't go around much.

What has Regis University to do with SANs?

The University was just awarded the Virtual Computing Laboratory Grant that, in the words of Dan Likarish lead faculty in the master's program at Regis: "is going to change the technology industry in Colorado".

In essence the grant gives Regis University a SAN that its students will be able to access online. What's exactly the grant? Read on.

The Colorado Institute of Technology (CIT), a Colorado state-funded organization whose mission is to improve technology employment in Colorado was awarded the equipment grant. CIT awarded the equipment to Regis University as well as an additional $50,000 to install the storage area network (SAN) and institute the program at Regis' technology department.

Hitachi Data Systems (HDS), a Fortune 100 manufacturer of data storage solutions, donated the $250,000 in equipment. Students will learn their skills operating and managing HDS storage hardware and software to gain practical experience.

The Storage Networking Institute (SNIA), has volunteered to host the SAN (Storage Area Network) equipment, which can be accessed anywhere in the world, at their location in Colorado Springs.

Regis University's Master's in Computer Information Technology department (MSCIT), is utilizing the equipment and instituting the program. Regis is currently the only University in the area that offers a certificate in SAN Networking.

Obviously great news for current and future students at Regis University, but let me whisper to the likes of EMC, IBM and others: Do you really want to let those students earn their SAN certificate on HDS stuff only?

The funny side of data protection

Two vendors, eVault and LiveVault, both offer online backup solutions, have come out almost at the same time with a rather humorous promotion.

Is this their subtle way to suggest that data protection is no laughing matter? Perhaps, but why don't you judge by yourself?

Start by visiting the LiveVault site where John Cleese (you may have seen his the previous clip ) has a new Webinar on data protection ready to roll.

Then, after collecting your thoughts, think of a possible turkey to share with the folks at eVault. No, not a bird, EVault wants a story worth telling on data protection, actually a "Turkey of a story". Check it out here, but keep in mind that this bird won't be pardoned.

Blogmania

How did we manage before we had blogs? I don't have an answer to that but may I suggest another place to visit? It's strictly storage related, you'll find no personal tales there.

Enter Hu Yoshida's blog where Hitachi Data Systems CTO, Hu Yoshida breaks his storage wisdom into ready to consume, very appetizing morsels.

Yes, You Can Benefit Today From 10-GbE In iSCSI Solutions

Okay Mario, you asked the question, so I'll attempt to respond.

Do you need 10G plumbing for your storage?

Yes, there are certainly immediate opportunities to deploy 10-GbE networking connectivity as part of iSCSI solutions. The big question of course is where can you justify the investment today - early in the declining price-point curve? One answer is Bandwidth Aggregation to the iSCSI target server.

Let's look at this from the perspective of how one provides sufficient bandwidth for iSCSI solutions today - multiple 1-GbE links. Let's say you're an organization that has 6-8 average servers each connected via 1-GbE (pretty normal). If you add an iSCSI target server into the mix to provision storage to all these servers (for all the right reasons) a single 1-GbE connection to that iSCSI Server is probably going to be over-subscribed. The solution - multiple 1-GbE links in the iSCSI server - 2,3,4 - whatever your capacity planning indicates. You configure the iSCSI traffic across these connections - perhaps dedicating one to the Exchange Server because it's a heavy user - putting the 3 low-activity servers across another - etc.

This approach works - but it's a more involved planning and configuration process - and it's something you have to optimize over time as server loads evolve. And, what if you have more servers and more load you need to support? At some point the complexity and overhead of the multiple 1-GbE NIC model becomes a problem.

We've demonstrated that our iSCSI target, WinTarget(TM), can process the kind of bandwidth available on a 10-GbE link. You can see those results here. Assuming your iSCSI target server has the disk subsystem horsepower to keep up with that level of traffic (controllers, disks, etc), how do you take advantage of that per-server capacity? Can you put enough 1-GbE NICs into that server to fully take advantage of the available services?

Take the same situation and apply 10-GbE connectivity. You deploy a 1-GbE / 10-GbE aggregation switch - say something like the ProCurve 3400cl, that provides 24 ports of 1-GbE and 2 10-GbE ports. That switch retails for around $10K (using optics, not the newer CX-4 copper 10-GbE connectivity, which will be less expensive). Install a single Neterion Xframe 10-GbE card in the WinTarget iSCSI Server and let the switch aggregate the 1-GbE server traffic through that single link. For fail-over, perhaps you put two 10-GbE Xframe cards in the WinTarget Server. Granted, those cards are about $4K each today, but the prices are coming down quickly - and when the CX-4 interface cards are available (and switches) - the cost will drop considerably.

So what would that get you? How about 7.5TB of storage available for either NAS or iSCSI services for somewhere in the neighborhood of $30K-$31K (street price, with two 10-GbE cards) - and 24 1-GbE ports to access it. With the availability of CX-4 interfaces - that price could quickly be <$25K.

For a company that has made the decision to upgrade their infrastructure with SAN capabilities - I'd argue that this is potentially a very attractive alternative to FC.

Obviously individual circumstances will decide whether a 10-GbE investment is warranted - but there are certainly situations where adding this to the mix will make an iSCSI SAN solution an attractive option.

Paging your blocks of data

Doesn't happen often, but some new products change dramatically the landscape of their technology. One good example is virtual memory, a feature present and given for granted today in any computer worth its silicon content. Try to imagine what life would be if computers had only real memory to run their applications.

Probably we would be spending gazillion of dollars on memory chips to make sure that our most memory-hungry applications could run.

Of course virtual memory freed us from babysitting RAM: in modern computers the OS keeps tracks of how much memory an application needs and will automatically swap pages of data or code between RAM and less expensive virtual memory.

The application and its users don't know and don't care where that page is, because virtual memory has made the actual location of a page of data irrelevant.

Why am I rambling about virtual memory? Because a new application from Compellent, Data Progression, promises to do for storage what virtual memory did for computing.

If you forgot about Compellent here is a link to a review of their SAN.

In essence, Data Progression makes possible moving infrequently referenced blocks of data to a different storage device, automatically and seamlessly.

To understand how this works, it's important to remember that a Compellent SAN has a rather unique architecture. For example, you can define LUNs for say 10 TB even if your SAN has actual disk space for only 1TB. As long as you don't use all the space immediately, this will work, and you can add more disk space gradually without breaking anything.

One more thing to keep in mind: When creating a LUN, Compellent will suggest to assign different tiers of storage to accommodate less frequently referenced data.

For example a fast array with 15K RPM, 36GB SCSI drives could be the primary tier for that volume, while a slower but more spacious array with 200 GB SATA drives could be the secondary tier.

You can create more than two tiers, and I believe Compellent will add other media besides disks in the future. Also, you can further discriminate where to store data according to the RAID level, for example, RAID 5 for primary, RAID 1 for other tier if this make sense with your application.

Interesting so far? Let's recap. A Compellent volume can spread across different tiers of storage, pretty much like a blanket can lay across the different steps of a stair. Great, but how does the system decides where to store data?

Another morsel of information will help understand that: Compellent keeps (again automatically) some interesting metadata for each block of data. For example, frequency of access, what tier of storage is it on, what RAID level, which volume, etc...

Time to bring all this together. Let's assume that a volume contains Office files and has tier 1 on RAID 5 SCSI and tier 2 on RAID 1 SATA.

It's intuitive that the most recently changed files will remain on the expensive SCSI drives on tier 1, while older stuff will in time migrate to the less expensive second tier.

What happens if user Jean creates a file and doesn't touch it for months? Without her knowledge, Compellent will move the blocks of that file to tier 2.

What should Jean do when she needs to update that file 10 months later? ? Absolutely nothing: Jean will be able to open her file without knowing that the system is pulling those blocks out of the secondary tier.

Will there be a delay? Compellent swears on no delays and although I haven't worked with Data Progression yet, my guess is that if there is an access time difference with more recent files Jean won't even notice.

What's the impact on your servers when using Data Progression? Again, nothing changes: the LUN is just another volume. The OS won't even know that there are several arrays or RAID levels behind that LUN.

So is this application from Compellent going to kill all the hype about ILM (information lifecycle management)?

Honestly I don't know, but Data Progression has little to do with ILM, in my view, because it lacks (at least for now) a main prerequisite: classifying data according to business criteria.

What Data Progression can do quite well, is to keep recent data on your most expensive tier of storage and automatically move old stuff somewhere else, without a dedicated babysitter and with minimal increase of administrative cost. In fact, the application starts at $10,000.

Perhaps the most likeable aspect of Data Progression is that it works without affecting users or applications, just like virtual memory.