SAN

HDS makes incremental updates to midrange disk arrays

Hitachi Data Systems’ (HDS) AMS 2000 series got a touching up today with the announcement of some incremental updates to the midrange disk array.

HDS is making two updates available now - a new High-Density Storage Expansion tray and a NEBS-certified DC power option for the 2500 model.

The High-Density Storage Expansion Tray holds up to 48 one-terabyte SATA disk drives in 4U; existing AMS trays hold 15 SAS or SATA drives in 3U. The maximum number of drives supported in the 2500 (480) hasn’t changed, but the maximum configuration now takes up one less rack than with the 15-drive trays. Good news for users focusing on storage and energy efficiency.  A fully loaded high-density tray is listed at $83,260.

The AMS 2000 series has had the option of running on battery power (DC) since the arrays were first announced last fall, but the new 2500DC model has been certified as compliant with the Network Equipment Building System (NEBS) standard for use in telecom and other “lights out” environments.

According to HDS senior product marketing manager Mark Adams, there’s little technical difference between the certified and non-certified versions, but the certified version “has been proven operational through intense earthquake activity” and certified by an independent lab. Another difference between the NEBS-certified and non-NEBS certified models is the price: the compliant list price is $102,870, while the non-compliant list price is $92,500. 

Later this year, HDS will make 8 Gbps Fibre Channel host ports available for the AMS 2300 and AMS 2500 models (internal disks will remain SAS or SATA). Security features to become available in the second half of 2009 include support for external authentication, meaning the AMS array and authenticating server don’t have to reside on the same network. Finally, as announced last week, HDS is extending its Dynamic Provisioning (HDP) software to run on the AMS in addition to the high-end USP-V.

User Matt Stroh, SAP business administrator for Wisconsin-based Industrial Electric Wire and Cable (IEWC), said he’s looking forward to deploying thin provisioning for the AMS 2300 he bought to replace an EMC Clariion CX-300 and HDS AMS 500 at the beginning of the year. “I’d like to get my hands on that as soon as possible,” he said. “We have a lot of file systems just storing SAP and Oracle binaries, and I don’t need much storage for them, but I’ve been giving them a big chunk anyway.”

While dynamic provisioning is going to be available for AMS, the Zero-Page Reclaim feature recently announced for the USP-V version of HDP will not be available for the forseeable future, according to HDS officials, who have not disclosed a technical reason why that’s the case.

3Par braces for V-Max

3Par CEO Dave Scott spent a lot of time on his company’s earnings call Tuesday evening talking about EMC’s new Symmetrix V-Max. That makes sense, considering 3Par probably has the most to lose of all EMC rivals if V-Max is a hit with customers.

The V-Max EMC launched a month ago is a nod in some ways to 3Par’s modular cluster-node architecture, and a move away from the giant monolith enterprise system. 3Par had success while people were waiting for the new Symmetrix – Tuesday it reported revenue of $48.5 million last quarter, an increase of 37% from last year and 1% from the fourth quarter of last year. That’s compared to declines in EMC’s Symmetrix and Clariion midrange systems of 18% year-over-year and 25% from the previous quarter. But what happens to 3Par’s InServ business if EMC’s sales spike from V-Max?

Scott came to his company’s earnings call prepared to talk about V-Max, and launched into a lengthy answer when asked about it. He laid out what he considers V-Max’s failings – no improved RAID management or ASIC-assisted workload, poor thin provisioning and limited support for wide striping, the system is an untested “version one” of a new architecture, and so on. “In other words,” Scott concluded, “it does not have the agility or efficiency necessary for utility computing and virtual data centers. It seems to have missed the mark in much the same was the [IBM] XIV did.”

Of course, EMC has already made its own case for the V-Max contradicting many of Scott’s points, and will continue to try and press its case to 3Par customers. One feature where EMC is unquestionably ahead is in its support of solid state drives (SSDs). 3Par is the last major storage system vendor to add SSDs to the mix, and Scott says it’s in no hurry to jump on the bandwagon.

“We believe that solid state disk will have a kind of meaningful place in data storage, but the price performance characteristics of it have to change,” he said. “You should expect to see us include solid state disk maybe around the turn of the year, but the major benefit that solid state disk provides is something we achieve through autonomic wide striping, which is not necessarily available to many of the legacy incumbents’ architectures. So our need for solid state disk is not nearly as significant as theirs.”

Pillar claims SPC-1 supremacy

Pillar Data Systems has published its first system test benchmarks for the Axiom 600 disk array via the Storage Performance Council (SPC). It tested a system with 42 total TB, including 10 TB used and mirrored to allocate a total of about 20 TB. The results were 64,992 total IOPS.

These numbers come in ahead of competitive systems from IBM, NetApp and EMC. The EMC CX 3-40 was tested last January by archrival NetApp, making its SPC benchmarks controversial, but as listed a 22 TB system with 8.5 TB used and mirrored produced 24,997 SPC-1 IOPS. A NetApp 3170 with 32 TB and 19.6 used and mirrored resulted in 60,515 June 10. A 37.5 TB IBM 5300 with 13.7 TB used and mirrored produced 58,158 SPC-1 IOPS Sept. 25.

I found it interesting that with one 2,000-IOPS deviation between the IBM 5300 and NetApp 3170, the systems generally performed better according to which had been most recently tested. Note also how much of an outlier EMC is, both in terms of capacity used and total capacity. It was also an outlier in its free space, with just under 1 TB unused. IBM and NetApp both left approximately 5 TB of free space in their configurations, and Pillar had 16 TB of free space.

I do have to wonder how much weight users give to these industry benchmarks when selecting a product.  NetApp’s submitting EMC systems to SPC, a flap last summer over server virtualization benchmark testing, and  continued inconsistency among vendors as to who submits systems for benchmarking leaves a lot of potential reasons to take benchmarks with a grain of salt.

Dell prepares for converged networks, FCoE and iSCSI

During a conference call with storage reporters today to discuss the future for data center networking, Dell senior storage manager Eric Endebrock pointed to the convergence of Ethernet and Fibre Channel as inevitable. ”Change is afoot,” he said. “FCoE is a more straightforward management infrastructure–the next generation of intercommunication for Fibre Channel.”

No suprise there. Practically every FC storage vendor is saying that. But where it gets tricky with Dell is, it dropped $1.4 billion in iSCSI SAN vendor EqualLogic less than a year ago. And where EqualLogic’s PS Series iSCSI SAN arrays fit into the converged picture isn’t clear yet.

“Protocols will not necessarily be the top factor in choosing the next storage system for customers,” Endebrock said. “We get caught up in the latest cool technology trend on [the vendor and press] side, but customers don’t necessarily care about that.” He added that lossless Ethernet “will float all storage boats” and that “customers see a place for all protocols.”

Also, “linking EqualLogic to iSCSI is probably not the best way to think about it–we also provide a scaling architecture and solve higher customer needs–it’s far more than just a protocol discussion.”

So far, Dell spokespeople aren’t willing to go into further detail about what its exact plan is for EqualLogic. “We continue to investigate our options and will support 10 Gigabit Ethernet as well as Data Center Ethernet with EqualLogic. We’re going to watch our customers’ needs and what the customers want,” Endebrock said.

A presentation at Storage Networking World titled “Yes, Fibre Channel and iSCSI Can Coexist” by director of global storage and network marketing Praveen Asthana, offered some clues about how Dell sees it all fitting together. “Mixed is in,” Dell’s Asthana said.  But he identified Ethernet as the glue–whether it’s providing the base layer of the unified network or providing a simple management and monitoring interface for all endpoints on an IP network.

While traditional Fibre Channel offers better performance for business applications than traditional iSCSI, it also offers better performance for streaming applications and high-performance computing (HPC) workloads, Asthana pointed out. But he also projected scale-out iSCSI, especially with 10 GbE, will surpass the performance offered by both earlier protocols.

FCoE is still  a topic that remains in the eye of the beholder. An attempt by FC vendors to stay relevant against 10 GbE? Or Ethernet taking over the data center? Depends on who you talk to.

Bottom line: Dell will support FC as long as it supports Clariion. Endebrock was mostly mum when it came to the relationship with EMC, as addressed by EMC CEO Joe Tucci in the company’s third-quarter earnings call on Wednesday. “Joe actually laid out that we have a great relationship and we’re actively working together on how to go to market on the best way possible, working on fitting our product lines together. We’re going back to basics and at the ground level refocusing on where we’ve seen success in the past.”

Marc Farley, citing allergies, bolts Dell/EqualLogic

The turnover that often occurs when a large company buys a smaller one has begun at Dell/EqualLogic, about seven months after Dell closed its $1.4 billion acquistion of the iSCSI SAN vendor. The most visible of these departures was delivered via vlog by Inside IT blogger and Dell/EQL evangelist Marc Farley, who gave his reason for leaving as “I’m not much of a big-company guy. I don’t know if I’m allergic to them, or what.” Farley said he’d be going to work for a smaller storage vendor, but didn’t specify which.

This is similar to what I heard recently from Roman Kichorowsky, former director of PR for EqualLogic, whose name is on a Dell press release as recently as July 10 but who has now moved on to FalconStor.

It’s not unusual for employees to leave in these kind of acquisitions. There are real cultural differences with almost every merger, and you’ll always find people who smile big until the press coverage dies down, wait for options to vest, and then get out.

But Dell’s retention of EqualLogic’s service and support staff as well as channel partners were major concerns for EqualLogic customers after the acquisition. The departure of PR and marketing people almost surely won’t affect the customers. But it’s worth keeping an eye on who else chooses to exit.

QLogic boss says he’s too busy to quit

 QLogic CEO H.K. Desai had planned to step down this year, but says there is too much going on in the storage industry for him to leave now.

That’s why his heir apparent left QLogic in March, and QLogic isn’t looking for a replacement. “I’ll be hanging around for awhile,” Desai, who has been QLogic’s CEO since 1995 and its chairman since 1999, told me.

He wouldn’t have said that a year ago. QLogic hired Jeff Benck from IBM as COO and President in April 2007 with the understanding that he would replace Desai in a year. But Desai stayed put and Benck left QLogic - eventually signing on with its HBA rival Emulex as COO.

Desai says QLogic isn’t trying to replace Benck, because it doesn’t need to replace the CEO. “The whole plan was for us to have a long-term transition, and it didn’t work out,” Desai said. “I’ve been fully engaged the last few months and I expect to be fully engaged for awhile. I won’t even talk to the board [about a COO] unless we plan a transition, which I don’t expect to do.”

Desai said his change of heart came because the storage networking world is so busy now. The industry is in a transition to 8-Gbps Fibre Channel with Fibre Channel over Ethernet (FCoE) and 16-gig FC coming up on its heels. QLogic is also trying to make inroads in InfiniBand. “There’s so much activity going on now, I don’t want to disturb anything,” he said. “Our team has been working together for so many years, and I don’t want to bring anybody in from the outside to disturb things. There’s a time to do a transition, and I don’t think this is the time.”

Of course, if Desai is looking for a lull in emerging storage technologies to retire he just might work forever.

Dell moves to top of growing iSCSI SAN market

IDC confirmed Dell’s claim that it has moved into the No. 1 spot in iSCSI SAN market share following its $1.4 billion acquisition of EqualLogic. According to the IDC quarterly storage numbers for the first quarter released today, Dell passed EMC and NetApp to take the lead with 27.7 percent of the iSCSI market. Overall, IDC pegged iSCSI as accounting for 6.1 percent of the overall external disk storage revenue — up from 4.1 percent a year ago and 5.1 percent in the fourth quarter of 2007.

It appears that a good chunk of the market share Dell gained came at the expense of its storage partner EMC. EMC, which co-markets Clariion systems with Dell, slipped from 18.2 percent to 15 percent of iSCSI market share in one quarter. Does that mean Dell customers are buying EqualLogic systems instead of Clariion AX iSCSI boxes? Maybe, although Dell still accounts for about one-third of overall Clariion sales.

NetApp increased its iSCSI share from 18 percent to 20.5 percent to move ahead of EMC while slipping behind Dell. No other major vendor has more than 4.7 percent of the iSCSI market, with the “other” (all vendors except for Dell, EMC, NetApp, HP, IBM, Hitachi, and Sun) category at 29.3 percent - nearly twice as much as “others” have of the Fibre Channel market. The others’ iSCSI share actually came down - it was 40.1 percent in the fourth quarter — reflecting  the shift of EqualLogic’s revenue from others to Dell.

It’s a good bet that LeftHand Networks sits in fourth place overall with a large piece of the others’ share. LeftHand was considered a close second to EqualLogic among private iSCSI vendors before Dell scooped up EqualLogic. LeftHand remains private and we don’t know its financials, but marketing VP Larry Cormier said it is picking up 200 to 300 customers a quarter. . .some of those because EqualLogic is no longer independent. “Some shops just won’t buy Dell,” he said.

In any case, with iSCSI drivers such as server virtualization and the eventual emergence of 10-Gig Ethernet fueling interest and the vendor landscape changing, the iSCSI space will be interesting to watch over the next year. . .just as converged Fibre Channel/Ethernet networks may blunt iSCSI’s encroachment into the enterprise.

Apple’s move away from hardware lock-in to low-cost generic arrays is a shrewd one

I’ve been seeing the scuttlebutt about Apple and Promise Technology and couldn’t help but add my two cents about how many Promise arrays I’ve seen pop up lately.

Last week, while installing our IBM N-series, I saw a couple of admins installing a multi-shelf Promise array. Peering through the cages in one of our colo areas, I’ve seen quite a few Promise and generic arrays installed. Walking the aisles in the areas I have access to, I’ve seen a rapid uptick in the installation of Off-Broadway-brand array vendors.

We own a small (5TB) Promise V-Track array we use for limited duty validation and testing (we bought it before the Storevault was released). I like it — it certainly fills the need and it does what it’s supposed to do. I can buy any brand and size SATA hard drive I want and the management tools come with the product at no additional charge. I was able to set it up in about 30 minutes and after the drive initialization (took close to 24! hours) I was all set and ready to go, all without a PhD. I even did the guy thing and didn’t read the instructions! I don’t know about you, but I can’t really ask for more, considering the price.

I’ve seen the folks at Apple accused of being stupid or lacking foresight in the past (Steve, I’m still upset about my Newton!!). In recent years, the accusers have usually been dining on crow, given the fact that Apple’s products consistently create trends. (Anyone up for an iDog?) I firmly believe they know something about the trend towards lower-cost generic arrays using generic disks in generic trays, otherwise (at least in my mind anyway), a company that prides itself on solidly locking you into their hardware when you use their software would have gone with a more mainstream storage vendor, or simply re-branded something and inserted a v-chip.

You’ve read me typing this for a couple of blog posts now, but I’ll type it again: Small to midsized SANs for under $50,000 with simple software and easy to use interfaces are going to be the market in the coming years. I’ll go a step further and say the days of proprietary drive trays and “enterprise-class ” drives are numbered too.

I seem to recall another big vendor that often gets maligned for lacking foresight snapping up a low-cost storage array vendor recently.

More importantly, Apple knows how to make difficult things easy and stylish. Not to mention that people who OEM for Apple (Foxconn , Acer et al.) are quite happy pumping out the iWhatever. It wouldn’t be too far-fetched to see Promise doing the same.

If there was ever a company that could pull off making a product that does easy data migration … see where I’m going with this?

Couple Apple’s really-easy-to-use SAN software with low-cost generic arrays and you could have a quick rise to major player in the storage software market. . .for a company many thought would be out of business by now, bringing in another company that “real” storage vendors look down their nose at.

Intel gets inside of FCoE

Fibre Channel vendors aren’t the only ones pushing the new Fibre Channel over Ethernet (FCoE) standard designed to help Fibre Channel devices take advantage of 10-gig Ethernet.  Intel is also getting into the game, with an FCoE Linux initiator.

Intel this week released an open source FCoE initiator that it will maintain on http://www.open-fcoe.org/. The FCoE initiator will work the way iSCSI initiators wok on current IP SANs. By going open source instead of developing the initiators for its own products, Intel hopes to accelerate the availability of FCoE by getting feedback from the Linux community. Intel storage planner and technologist Jordan Plawner said the goal is for Linux servers to ship FCoE-ready, just as they ship with iSCSI inititators today.  

“We believe 10-gig Ethernet provides an opportunity to converge SAN and LAN traffic,” Plawner said. “We’ll continue to support iSCSI, but FCoE makes it easier to connect Ethernet into Fibre Channel SANs.”

That’s the party line for Fibre Channel vendors, and one that iSCSI SAN proponents dispute. Like iSCSI vendors, Intel is looking at it from the Ethernet side - but Plawner said FCoE will be better suited than iSCSI to take advantage of the coming Enhanced Ethernet spec. Enhanced Ethernet is a new version in the works that boosts Ethernet’s performance to make it more suitable to run storage.

“It’s much easier to adopt FCoE for Enhanced Ethernet,” Plawner said. “iSCSI is Ethernet end to end, so you would need a completely new subnet because you need Enhanced Ethernet on every node. With FCoE, just the first server and first top-of-the-rack-switch needs Enhanced Ethernet.”

Intel is looking to put FCoE Linux initiators on adapter cards that will work with FCoE switches in 2008. Plawner says he expects FCoE-enabled switches from Brocade and Cisco in the second half of next year, and he thinks companies will deploy FCoE in their networks by the end of 2008.

Plawner’s time frame is even more optimistic than that of some Fibre Channel storage vendors backing FCoE. Brocade execs says they don’t expect adoption until 2009, and they don’t think widescale adoption will arrive before 2010. But Brocade pledges to support FCoE in the DCX backbone director it will launch next year. Cisco’s FCoE switches are expected from Nuova Systems, which is 80 percent owned by Cisco. Nuova has yet to give product details, but industry sources say it will likely have FCoE switches or cards that plug into Cisco MDS switches early next year.

Here is a more detailed explanation of how Fibre Channel and Ethernet can converge.

Buying typical storage for video surveillance? Rethink that!

Up until now you (corporate IT) have not had to worry about video surveillance. That job was up to the security guys, those guys that wore uniforms and pretty much kept to themselves. But be prepared. If you are not already deeply involved in video surveillance equipment RFP creation, acquisition, installation and management, you will be very soon.

The world of video surveillance is changing so rapidly that the user and the traditional supplier are both in a state of frenzy. It is within this transformation that the role of IT is becoming increasingly critical. The reasons for the increase in video surveillance are pretty easy to understand. Post 9/11, enterprises as well as governments are all adding or increasing video surveillance to the security equation. Of course, casinos and banks have always been the leading users of video surveillance, but now everyone is in the game. On a typical day, a person living in a city may be videotaped five or more places, as he drives to work (and passes through specific traffic lights), parks his car in the company parking lot, enters the building, makes a trip to the bank at lunch, grabs a couple of items at the local K-mart and heads home. There are all kinds of privacy issues that can be debated, but I am staying away from that. At least for now. Right now, I am more interested in the technology and IT’s increasing role in video surveillance.

Traditional video surveillance equipment was not designed to deal with this onslaught and is gasping for air. It is being replaced almost completely with IP-based equipment.  That’s where you come in. Until now, most video surveillance equipment was based on CCTV (closed circuit TV), which basically meant the cameras, which recorded analog video, were hooked up via coaxial cable to the central point, where the video was taped on VCRs. Later, DVRs converted the analog signal to digital at the central location before storing it. But, these technologies cannot deal with the onslaught of data from more and more cameras and the fact that cameras are increasingly adding higher resolutions.

The latest crop of cameras records video in a digital format, and compresses it using MJPEG or MPEG before transmitting it over standard IP network to a central location that stores the data on scalable disk arrays. Once in the realm of IP, all the goodies we are used to in IT become available to an industry that still thinks of guards manning physical structures. Centralized management become feasible, data can be accessed asymmetrically, from multiple locations, replicated when appropriate. Another level of sophistication is being added at the end points. Now cameras can be activated when they detect motion or switch into a higher resolution if certain criteria are met. Video analytics allow software to recognize facial characteristics. Searches can be conducted for specific objects or people. You get the idea. It is like James Bond gadgetry becoming available to regular folks. But, that is reflective of the world we live in.

I think you (IT) need to be prepared to play a major role in this transformation that is occurring. You are the resident experts in storage and, at this point, pretty well up on IP technologies as well. Video surveillance simply becomes another application you have to support. So, if you are not already deeply involved in the selection and day to day management of the video surveillance equipment, it is only a matter of time. Security people who used to make decisions on such purchases without any consultation, will now insist on your involvement. You should gladly offer to help.

Another important thing to realize is that the type of storage you end up selecting for these applications will very likely be different than storage for other applications. For video surveillance the attributes that matter for storage include cost effectiveness (dirt cheap), highly scalable across both capacity and performance (cannot afford to create islands of storage), low entry price point, cost effective availability (mirroring may be too expensive), protection from disk drive or nodal failure and, most importantly, it needs to IP-based. Everything else in this environment is IP based, so making storage IP-based makes it easier to understand and manage. FC storage would bring in a level of complexity that is unnecessary here. Also, legacy architectures that have grafted an IP (iSCSI) interface would not cut it here, because they would not meet the other requirements above. Storage players that I believe merit consideration include Pivot3, Intransa, LeftHand Networks and to a lesser degree, EqualLogic (their price point may be too high for this application). There are other inexpensive storage offerings, such as from Nexsan or Xyratex but if an architecture does not allow clustering and presentation of a single system image, as it scales, it misses a criterion that I consider absolutely necessary for this application. However, you may want them in the initial mix as you start the evaluation process. I am sure you have enough on your plate without adding yet another storage-hungry application. But the way the winds are blowing, you either pro-actively plan on this or you will get pulled in pushing and screaming.