StorageMojo

The window for Blu-ray success is rapidly closing (see Blu-ray is dead. Heckuva job, Sony!). Which means that 50 GB writable disks will never cost $0.35 a piece.

What is a storage hungry consumer to do?

Massive removable/transportable storage
Together cheap CD/DVD media, thumb drives and ever-growing file sizes killed floppies, Zip drives and all the other removable magnetic disk media. Removable optical media may be next.

Historically, successful PC removable media have stayed in a fairly narrow capacity band relative to hard drives - somewhere between 10x and 50x. If the average PC has a 500 GB hard drive then a removable media between 10 GB and 50 GB is needed.

Dual-layer DVDs are just on the ragged edge of that number while dual layer Blu-ray could handle hard drives up to 2.5 TB. If Blu-ray will never achieve the ubiquity and low cost of DVDs what will fill the gap?

Meet the candidates
Flash drives are a promising alternative. Large capacity thumb drives are available today for about $2/GB. In 2 years those drives will be $.50 a gigabyte or less.

That doesn’t really compare to a Blu-ray writable media at $2 for 50 GB in 2 years, if that comes to pass. But flash drives do have advantages in size weight, ruggedness, and the ubiquity of USB ports.

Cloud storage is another option. For example, Mediasilo offers a service that stores and password protects individual files so the owner can control distribution.

Network bandwidth is the key bottleneck. Countries that have much faster Internet access than the United States could put this to work with some fairly large files. Here in the USA however it won’t be practical for large files for years to come.

Disk drives may be the most promising option. Smaller drives with better shock specs - check out Toshiba’s new 250 GB 1.8 inch drive - could handle just about any transportable application.

If drive vendors would start specifying their drives as an archive medium - even if it’s just five years - they might be able to sew up the home and SOHO archive market. The iVDR initiative is a step in the right direction, but like Blu-ray they are trying to serve too many markets.

Mempile and Inphase, 2 different approaches to high-capacity optical storage, are unlikely successors. Both appear viable for niche markets, but massive investment is required to take them to consumer ubiquity. What app will take them there?

The StorageMojo take
Removable media are the fruit flies of the storage industry: fast breeding and short lives. The pace of innovation is a good thing but a consumer standard needs a 10 year life.

Letting Hollywood drive next-generation media formats doesn’t work. Downloading will supplant physical media for consumers.

There will be no optical replacement for the writable DVD that offers its ubiquity, low cost and interchangeability. Like removable consumer magnetic media, optical has reached a dead end because there is no consumer driver for 250 GB+ optical media.

The good news: consumers will want to archive. There is a huge opportunity for the company that can figure it out.

Courteous comments welcome, of course.

Every silver lining has a cloud
As Storage Networking World’s go this was quiet. It is always good to get together with industry colleagues. And to be able to look executives in the eye when I asked them hard questions.

Everybody is antsy about the economy. With the IPO market shut down and the VC’s pulling back, now is not the time to be an undercapitalized or overspending startup. Horde cash and market wisely.

It is also not a good time to be flogging the Same Old, Same Old overpriced kit. With layoff rates rising fast and American underemployment at its highest level in 15 years “change” is in the air.

IT pros should be asking themselves “am I part of the solution, or part of the problem?” Because the gimlet-eyed finance guys are looking at you and asking the same thing.

Good news
The good news: there are options today that weren’t available during the dot-bomb meltdown.

The bad news: you’ll have to do some homework and talk to some new vendors to figure out what these companies can do for you.

StorageMojo’s favorite companies at SNW

• Hifn I have had a hard time getting a handle on this company - starting with pronouncing the name - but I’ve got it now: a storage network appliance that encrypts, compresses and, in a newly announced feature, de-duplicates. They don’t sell direct to end-users but maybe your integrator or SAN vendor resells them.
• Xiotech CEO Casey Powell announced in double digit quarter over quarter growth figures due to the success of their Integrated Storage Element product. ISE is a game changer for the storage industry, but Xiotech needs to do a better job of articulating the benefits to non-IT execs.
• Storewize offers an appliance that sits between clients and file servers. It compresses and deduplicates data before it gets to the filer. Not only does it stretch storage capacity but it also improves filer performance by reducing bandwidth requirements. Works at GigE wirespeed.
• Permabit has been shipping their cluster-based Enterprise Archive for two quarters. Permabit’s CTO, Jered Floyd, has a blog with a great post on why - among other things - it is time to stop talking about content-addressable storage (CAS).

However, my favorite this SNW is Axxana (see Axxana fixes the speed of light). They are a game-changer for high-end disaster protection. Too bad you can’t buy them today.

The StorageMojo take
Compared to the dot-bomb fiasco, storage companies are in better shape today. 8 years ago over-funded startups were buying big Symms and E10000s by the truckload. When the music stopped a lot of new kit was selling on Ebay for pennies on the dollar.

Endemic over-buying is not today’s problem. As the world-wide deleveraging continues the push to do more for less will accelerate.

Somebody is going to have to unravel the underlying value of all those toxic credit default swaps and securitized mortgages. That will take a lot of data storage.

The industry will take a hit, no doubt. But you won’t see the kind of dot-bomb u-turn that saw EMC sales drop several billion in one year.

The world’s financial industry, whose growth has exceeded worldwide GDP growth for the last 30 years, is in for a long spell of below average growth. Health care looks like a winner though.

Courteous comments welcome, of course.

Or a reasonable facsimile thereof
If you are interested in Disaster Recovery check out Axxana. They solve the limited synchronous data copy distance problem with a black box designed for data. Concept is simple but getting the details right is hard.

The problem
Synchronous replication requires that apps wait until the remote site completes the write. Given the speed of light, that means that synch sites can’t be very far away. Certainly not the 300 miles the SEC would like to see for financial institutions - we still have a few of those, don’t we?

Axxana’s answer
No matter what happens in a plane crash, they always seem to be able to recover the “black box” that tells them what the plane was doing shortly before the crash. Axxana has developed a black box for data centers.

Here’s how they describe it:

The Phoenix Black Box is located near the storage system at the primary data center and records a synchronous data stream from the storage. At the same time, an asynchronous data replication system is moving data to a secondary data center (the remote recovery site). The Phoenix Black Box has to protect only the Gigabytes of data that would have been lost in a typical asynchronous replication scenario. Data is protected inside the Black Box during the course of the disaster and can be immediately extracted.

Data extraction is achieved either by:

• Physically locating the system by tracking the homing signal and connecting a laptop with an Axxana software component to the Phoenix System™ at the disaster site, or
• The self sufficient and well protected system transferring the data to the secondary site using highly resilient cellular broadband technology.

Your data phones home after a disaster.

Compelling economics
It will take a while to suss out all the implications, but one simple scenario is a company with 3 data centers around the world could in-source their DR strategy with the equivalent of synchronous data recovery. How much would that save?

Distribution
They are working with as many of the major vendors as they can to get the product to you through people you already deal with. Expect to see some announcements.

The StorageMojo take
They are in contention for StorageMojo’s “coolest new product as SNW” award. It looks like they can handle anything up to an A-bomb blast. If that happens even synchronous data replication may not work. Besides, a dirty bomb is much more likely. Happy thoughts, eh?

Comments welcome, of course. Guys, sorry if I jumped the gun. But when I saw the web site was up . . . .

StorageMojo has decamped to Dallas for Storage Networking World this week. I’ll be leaving Wednesday afternoon.

I still look like my picture to your right. If you see me feel free to say hello. Always interested in hearing user stories about products and learning about new products.

The StorageMojo take
I like SNW. It has the highest concentration of storage questions and answers of any event I’ve found.

StorageMojo gets questions from baffled civilians out in prospect-land. This one seems worthy of a thorough airing.

The writer is a student and a storage newbie, but she has the kind of question that more folks are asking. Here’s her note, edited for clarity:

I am working on an archive. The main idea is to store large files of 90 Gb each. This archive is meant for research purposes (that’s why the files are so big) so we can think of storing millions of these files (20-40 million). I read the article about google’s warehouse and I think that’s the scale of the project.

I’m sending you an image [ed. note: a BlueArc system] with the kind of rack that I found so far (1TB/blade) but I am not completely sure that’s the latest technology I can use for this huge scale project.

By the other hand I have the problem of trying to figure out the electricity required for it (thinking of thousands of racks with blades working in a net with backup servers).

I appreciate your time and help in advance.

Let’s do the math
OK, arithmetic. 20 million 90 GB files is, um-m, 1.8 billion GB or, hm-m, 1.8 million TB, which is 1800 petabytes or 1.8 exabytes.

2 million 1.5 TB drives would just handle this with no redundancy. At an average price of $100 per drive - they are higher now of course, but unboxing 2 million drives would take some time - our student is looking at $200 million just for drives.

High density packaging is required to reach 500 3.5″ disks in a rack. The Sun Fire 4540 puts 48 drive, 4U box could support 480 drives in 42U - if OSHA would lighten up and you didn’t mind using a ladder to replace failed drives. Any other ideas?

The 4540 also supplies the server cycles this project needs. Other server/storage enclosures would have a hard time equalling the 4540’s density (note: 10 4540’s in a rack is illegal in the US and not practical anywhere) but HP’s ExDS9100 comes close.

At 750 TB per rack (500 drives), 4,000 racks could handle it. A cheap 4 post/42U rack might cost $200 in volume, so call it $800,000 for racks. I don’t know what the boxes would cost but it would be at least several times the racks.

This could also be a great application for Copan’s enterprise MAID systems if the workload is right. With up to 896 TB per rack they have the density. Price/performance would be the next issue to look at.

Would you like a data center with that?
4,000 racks require about 40,000 square feet of floor space. Racks typically account for about 1/3rd of a data centers floor space - hallways and other equipment occupy the rest - a 120,000 sq. ft. or 11,100 sq meter facility is required.

At $500/sq. ft. the data center would cost an additional $60 million. YMMV.

Oh, and redundancy is extra.

Power
The drive’s operational power consumption is 10 watts. Drives alone eat 20 megawatts. Side note: disk power consumption is the reason that storage vendors find differentiation on power consumption elusive.

Assume 1 250W server for every 100 drives means another 5 MW for servers - a low-side estimate. Leaving aside network infrastructure and lighting, the HVAC load for 25 MW is around 12.5 MW, according to some rules of thumb.

Let’s round up and call it 40 megawatts. You’ll want to locate this facility near the Columbia River to get cheap hydropower - maybe next door to Google in The Dalles, Oregon.

I haven’t deciphered BPA power pricing, but I’d guess 40 MW would run about $2 million a month. Copan, much less.

The StorageMojo take
Well, what do you know? Building an exabyte data center is feasible. All it takes is money - $400 million with all the goodies - and power.

Time to readjust the mental model of storage possibility. Other than the NSA’s acres of disk at Fort Meade though, I’m not aware of any exabyte data centers.

Technically, I’m not aware of the NSA’s facility either.

Google isn’t a good hardware model, since their storage density is only about 120 drives per rack, not 500. Plus they don’t use the highest capacity drives in favor of the lowest $/GB - drives nearing the end of their product life cycle. Google’s GB/sq. ft. is way lower than my interlocutor could afford.

Their software - BigTable and GFS - looks to be a good fit. But even Google doesn’t run 20,000 node clusters. That is Lustre territory.

Who has the best solution to the exabyte problem? Nominations are now open.

I remember hawking 1.8 - one point eight! - gigabytes in a single 42″ rack. For $50,000! A terabyte was almost beyond imagination. Today my home system has 2.5 TB. Whoa.

Comments welcome, of course. Who else could use an exabyte of data? Disclosure: I’ve done work for HP and Sun. None for the NSA, though.

Hewlett-Packard’s acquisition of the LeftHand Networks shows how cluster storage is going mainstream - and how HP plans to be right in the middle of it. First PolyServe and now LeftHand.

This is about commodity-based clusters
Not iSCSI or GigE or 10 GigE as a storage interconnect. Fibre Channel’s failure to move downmarket - and Infiniband’s similar problem - means GigE is the only game in town.

Reaching the huge, not currently imploding, SMB market requires meeting people where they live. SMBs don’t live in Fibre Channel glass houses. GigE isn’t ideal, but it’s cheap and it works.

Did HP overpay?
$360 million isn’t pocket change, but it is only about 4x the $86 million investors put in. The investors get some nice coin, but it isn’t the 10-bagger they were hoping for.

Once the Lefties go through the interminable internal HP meat grinder, sales will grow rapidly. I suspect they weren’t up to Isilon’s $100M in sales - maybe $70M - but LeftHand was much closer to profitability. Net net: the price looks fair for a market leader in a high-growth market.

HP vs EMC
Battle of the competing cluster storage visions. Polyserve handles files; LeftHand blocks. EMC’s Maui is aimed at large-scale distributed file storage, a utility that ISP’s might resell to SMBs, but nothing an SMB would implement on their own.

Which will win - and there’s room for both - rests on the answer to the question Are there economies of scale in storage?. If there are, small-scale clusters sales will suffer and Maui should win.

The StorageMojo take
This is cluster storage market skirmishing, not a pitched battle. That will come but right now everyone is feeling their way, coming into the market from different directions, waiting to see what clicks.

Right now though, HP seems to have the strongest position. XIV is too new; Maui even newer; Lustre too complex; Isilon is digging out of a big hole. HP has the pole position with implementable products today and the services to back them up. Should be a powerful combination.

Courteous comments welcome, of course. Disclosure: I’ve done some work for HP, Isilon and Sun.

I’ve been waiting a long time for the transition to 2.5″ drives. Now it appears that the wait is over and it is truly underway. The dynamics aren’t what I expected.

The impetus: installing a new WD 300 GB VelociRaptor 10k drive in my Mac Pro. It is a 3.5″ heatsink wrapped around a tiny 10k 2.5″ drive. I’ll be running some tests on it shortly.

Fujitsu’s 2.5″ bet
Fujitsu’s Don Jeanette tried to brief me last week on their new drives, but I was more interested in their experience with the 2 form factors. Although the company is off the radar in most disk discussions - stealth marketing never dies! - I remember back in the ’80s when their 9″ Eagle was the drive to beat.

Fujitsu exited the consumer 3.5″ market in 2001, but still offer 3.5″ enterprise drives along with 2.5″ small form factor (SFF) enterprise and consumer drives. Now half their enterprise drives are SFF, a market where they are 2nd only to Seagate.

The consumer SFF
Notebooks dominate consumer SFF uptake, but as I look at my huge Mac Pro enclosure, it is obvious that smaller tower systems would also benefit. In the space used for 4 3.5″ drives, Apple could plug 20 SFF drives - and increase capacity by 50%.

With efficient many core CPUs, SFF drives, higher density DRAM, Intel’s new QuickPath (pdf) architecture, smaller fans and plenums, a much more powerful system than an 8-core Mac Pro could be built into a box 2/3rds the size. Hey, and lose the power-hungry FB-DIMMs while you’re at it!

The enterprise SFF
Consumers drive volume, but the enterprise drives margin. I assume Xiotech is well along on designing enterprise-ready drive shelves.

The interconnect issues must be interesting: internal SAS switches or cheaper high fan-out configurations? No issues we didn’t see in the move from 5.25″ to 3.5″, but it is a chance for a clean sheet design.

I’d welcome a call from Xyratex to discuss the topic.

The Seagate rumor
I’ve heard from a couple of sources that Seagate is currently designing its last family of 3.5″ drives. They’ve had good success with the Savvio SFF enterprise drives, so it is now up to the OEMs to make the move.

The StorageMojo take
With Fujitsu as a credible 2nd source, OEMs have no reason not to start down the SFF path. While the nearline guys will be the last to move off 3.5″ drives, the combination of more spindles/IOPS/bandwidth per U/$/watt should be persuasive - especially against surging flash drives.

The revenue crossover could be 2010, while Fujitsu’s Jeanette predicted unit crossover in 2011. With 6 SFF vendors competing for the business the pace of innovation should accelerate - to the lasting benefit of all storage consumers.

Update: Belatedly came across an article from the October 3rd WSJ. The money points:

CHIBA, Japan — Japanese electronics company Toshiba Corp. would be interested in Fujitsu Ltd.’s hard-disk-drive business, if approached, an executive said Thursday. . . .

The Fujitsu spokesman declined to comment on whether it is in talks with Toshiba.

Fujitsu expects its hard-disk-drive operation will break-even or post a slight loss on estimated revenue of 370 billion yen ($3.49 billion) for its fiscal year ending March 2009.

While Toshiba won’t disclose the performance of its hard-disk-drive business, Mr. Izumi estimates it could generate about 10 billion yen in operating profit and 360 billion yen in revenue for the current fiscal year.

End update.

Update II: Pete Steege of Seagate commented with a couple of links to his blog on the topic. Following those I found some more info on 2.5″ arrays that I’d missed.

• Xyratex the enclosure maker (not, oops, Xiotech - sorry) has a 2.5″ enclosure, the OneStore SP1224s, for sale.
• The Dell MD1120 is a 2.5″ RAID subsystem that appears to be based on the Xyratex enclosure.

End update II.

Courteous comments welcome, of course.