“In times of universal deceit, telling the truth will be a revolutionary act.”
-- George Orwell
Well, it has been over two years since I first covered IBM's acquisition of the XIV company. Amazingly, I still see a lot of misperceptions out in the blogosphere, especially those regarding double drive failures for the XIV storage system. Despite various attempts to [explain XIV resiliency] and to [dispel the rumors], there are still competitors making stuff up, putting fear, uncertainty and doubt into the minds of prospective XIV clients.
Clients love the IBM XIV storage system! In this economy, companies are not stupid. Before buying any enterprise-class disk system, they ask the tough questions, run evaluation tests, and all the other due diligence often referred to as "kicking the tires". Here is what some IBM clients have said about their XIV systems:
“3-5 minutes vs. 8-10 hours rebuild time...”
-- satisfied XIV client
“...we tested an entire module failure - all data is re-distributed in under 6 hours...only 3-5% performance degradation during rebuild...”
-- excited XIV client
“Not only did XIV meet our expectations, it greatly exceeded them...”
-- delighted XIV client
In this blog post, I hope to set the record straight. It is not my intent to embarrass anyone in particular, so instead will focus on a fact-based approach.
- Fact: IBM has sold THOUSANDS of XIV systems
XIV is "proven" technology with thousands of XIV systems in company data centers. And by systems, I mean full disk systems with 6 to 15 modules in a single rack, twelve drives per module. That equates to hundreds of thousands of disk drives in production TODAY, comparable to the number of disk drives studied by [Google], and [Carnegie Mellon University] that I discussed in my blog post [Fleet Cars and Skin Cells].
- Fact: To date, no customer has lost data as a result of a Double Drive Failure on XIV storage system
This has always been true, both when XIV was a stand-alone company and since the IBM acquisition two years ago. When examining the resilience of an array to any single or multiple component failures, it's important to understand the architecture and the design of the system and not assume all systems are alike. At it's core, XIV is a grid-based storage system. IBM XIV does not use traditional RAID-5 or RAID-10 method, but instead data is distributed across loosely connected data modules which act as independent building blocks. XIV divides each LUN into 1MB "chunks", and stores two copies of each chunk on separate drives in separate modules. We call this "RAID-X".
Spreading all the data across many drives is not unique to XIV. Many disk systems, including EMC CLARiiON-based V-Max, HP EVA, and Hitachi Data Systems (HDS) USP-V, allow customers to get XIV-like performance by spreading LUNs across multiple RAID ranks. This is known in the industry as "wide-striping". Some vendors use the terms "metavolumes" or "extent pools" to refer to their implementations of wide-striping. Clients have coined their own phrases, such as "stripes across stripes", "plaid stripes", or "RAID 500". It is highly unlikely that an XIV will experience a double drive failure that ultimately requires recovery of files or LUNs, and is substantially less vulnerable to data loss than an EVA, USP-V or V-Max configured in RAID-5. Fellow blogger Keith Stevenson (IBM) compared XIV's RAID-X design to other forms of RAID in his post [RAID in the 21st Centure].
- Fact: IBM XIV is designed to minimize the likelihood and impact of a double drive failure
The independent failure of two drives is a rare occurrence. More data has been lost from hash collisions on EMC Centera than from double drive failures on XIV, and hash collisions are also very rare. While the published worst-case time to re-protect from a 1TB drive failure for a fully-configured XIV is 30 minutes, field experience shows XIV regaining full redundancy on average in 12 minutes. That is 40 times less likely than a typical 8-10 hour window for a RAID-5 configuration.
A lot of bad things can happen in those 8-10 hours of traditional RAID rebuild. Performance can be seriously degraded. Other components may be affected, as they share cache, connected to the same backplane or bus, or co-dependent in some other manner. An engineer supporting the customer onsite during a RAID-5 rebuild might pull the wrong drive, thereby causing a double drive failure they were hoping to avoid. Having IBM XIV rebuild in only a few minutes addresses this "human factor".
In his post [XIV drive management], fellow blogger Jim Kelly (IBM) covers a variety of reasons why storage admins feel double drive failures are more than just random chance. XIV avoids load stress normally associated with traditional RAID rebuild by evenly spreading out the workload across all drives. This is known in the industry as "wear-leveling". When the first drive fails, the recovery is spread across the remaining 179 drives, so that each drive only processes about 1 percent of the data. The [Ultrastar A7K1000] 1TB SATA disk drives that IBM uses from HGST have specified 1.2 million hours mean-time-between-failures [MTBF] would average about one drive failing every nine months in a 180-drive XIV system. However, field experience shows that an XIV system will experience, on average, one drive failure per 13 months, comparable to what companies experience with more robust Fibre Channel drives. That's innovative XIV wear-leveling at work!
- Fact: In the highly unlikely event that a DDF were to occur, you will have full read/write access to nearly all of your data on the XIV, all but a few GB.
Even though it has NEVER happened in the field, some clients and prospects are curious what a double drive failure on an XIV would look like. First, a critical alert message would be sent to both the client and IBM, and a "union list" is generated, identifying all the chunks in common. The worst case on a 15-module XIV fully loaded with 79TB data is approximately 9000 chunks, or 9GB of data. The remaining 78.991 TB of unaffected data are fully accessible for read or write. Any I/O requests for the chunks in the "union list" will have no response yet, so there is no way for host applications to access outdated information or cause any corruption.
(One blogger compared losing data on the XIV to drilling a hole through the phone book. Mathematically, the drill bit would be only 1/16th of an inch, or 1.60 millimeters for you folks outside the USA. Enough to knock out perhaps one character from a name or phone number on each page. If you have ever seen an actor in the movies look up a phone number in a telephone booth then yank out a page from the phone book, the XIV equivalent would be cutting out 1/8th of a page from an 1100 page phone book. In both cases, all of the rest of the unaffected information is full accessible, and it is easy to identify which information is missing.)
If the second drive failed several minutes after the first drive, the process for full redundancy is already well under way. This means the union list is considerably shorter or completely empty, and substantially fewer chunks are impacted. Contrast this with RAID-5, where being 99 percent complete on the rebuild when the second drive fails is just as catastrophic as having both drives fail simultaneously.
- Fact: After a DDF event, the files on these few GB can be identified for recovery.
Once IBM receives notification of a critical event, an IBM engineer immediately connects to the XIV using remote service support method. There is no need to send someone physically onsite, the repair actions can be done remotely. The IBM engineer has tools from HGST to recover, in most cases, all of the data.
Any "union" chunk that the HGST tools are unable to recover will be set to "media error" mode. The IBM engineer can provide the client a list of the XIV LUNs and LBAs that are on the "media error" list. From this list, the client can determine which hosts these LUNs are attached to, and run file scan utility to the file systems that these LUNs represent. Files that get a media error during this scan will be listed as needing recovery. A chunk could contain several small files, or the chunk could be just part of a large file. To minimize time, the scans and recoveries can all be prioritized and performed in parallel across host systems zoned to these LUNs.
As with any file or volume recovery, keep in mind that these might be part of a larger consistency group, and that your recovery procedures should make sense for the applications involved. In any case, you are probably going to be up-and-running in less time with XIV than recovery from a RAID-5 double failure would take, and certainly nowhere near "beyond repair" that other vendors might have you believe.
- Fact: This does not mean you can eliminate all Disaster Recovery planning!
To put this in perspective, you are more likely to lose XIV data from an earthquake, hurricane, fire or flood than from a double drive failure. As with any unlikely disaster, it is best to have a disaster recovery plan than to hope it never happens. All disk systems that sit on a single datacenter floor are vulnerable to such disasters.
For mission-critical applications, IBM recommends using disk mirroring capability. IBM XIV storage system offers synchronous and asynchronous mirroring natively, both included at no additional charge.
For more about IBM XIV reliability, read this whitepaper [IBM XIV© Storage System: Reliability Reinvented]. To find out why so many clients LOVE their XIV, contact your local IBM storage sales rep or IBM Business Partner.
technorati tags: IBM, XIV, DDF, RAID-5, RAID-10, RAID-X, RAID-6, RAID-DP, HP, EVA, HDS, USP-V, EMC, CLARiiON, V-Max, Disaster Recovery, HGST, UltraStar, A7K1000
Did IBM XIV force EMC's hand to announce VMAXe? Let's take a stroll down memory lane.
In 2008, IBM XIV showed the world that it could ship a Tier-1, high-end, enterprise-class system using commodity parts. Technically, prior to its acquisition by IBM, the XIV team had boxes out in production since 2005. EMC incorrectly argued this announcement meant the death of the IBM DS8000. Just because EMC was unable to figure out how to have more than one high-end disk product, doesn't mean IBM or other storage vendors were equally challenged. Both IBM XIV and DS8000 are Tier-1, high-end, enterprise-class storage systems, as are the IBM N series N7900 and the IBM Scale-Out Network Attached Storage (SONAS).
In April 2009, EMC followed IBM's lead with their own V-Max system, based on Symmetrix Engenuity code, but on commodity x86 processors. Nobody at EMC suggested that the V-Max meant the death of their other Symmetrix box, the DMX-4, which means that EMC proved to themselves that a storage vendor could offer multiple high-end disk systems. Hitachi Data Systems (HDS) would later offer the VSP, which also includes some commodity hardware as well.
In July 2009, analysts at International Technology Group published their TCO findings that IBM XIV was 63 percent less expensive than EMC V-Max, in a whitepaper titled [COST/BENEFIT CASE
FOR IBM XIV STORAGE SYSTEM Comparing Costs for IBM XIV and EMC V-Max Systems]. Not surprisingly, EMC cried foul, feeling that EMC V-Max had not yet been successful in the field, it was too soon to compare newly minted EMC gear with a mature product like XIV that had been in production accounts for several years. Big companies like to wait for "Generation 1" of any new product to mature a bit before they purchase.
To compete against IBM XIV's very low TCO, EMC was forced to either deeply discount their Symmetrix, or counter-offer with lower-cost CLARiiON, their midrange disk offering. An ex-EMCer that now works for IBM on the XIV sales team put it in EMC terms -- "the IBM XIV provides a Symmetrix-like product at CLARiiON-like prices."
(Note: Somewhere in 2010, EMC dropped the hyphen, changing the name from V-Max to VMAX. I didn't see this formally announced anywhere, but it seems that the new spelling is the officially correct usage. A common marketing rule is that you should only rename failed products, so perhaps dropping the hyphen was EMC's way of preventing people from searching older reviews of the V-Max product.)
This month, IBM introduced the IBM XIV Gen3 model 114. The analysts at ITG updated their analysis, as there are now more customers that have either or both products, to provide a more thorough comparison. Their latest whitepaper, titled [Cost/Benefit Case for IBM XIV Systems: Comparing Cost
Structures for IBM XIV and EMC VMAX Systems], shows that IBM maintains its substantial cost savings advantage, representing 69 percent less Total Cost of Ownership (TCO) than EMC, on average, over the course of three years.
In response, EMC announced its new VMAXe, following the naming convention EMC established for VNX and VNXe. Customers cannot upgrade VNXe to VNX, nor VMAXe to VMAX, so at least EMC was consistent in that regard. Like the IBM XIV and XIV Gen3, the new EMC VMAXe eliminated "unnecessary distractions" like CKD volumes and FICON attachment needed for the IBM z/OS operating system on IBM System z mainframes. Fellow blogger Barry Burke from EMC explains everything about the VMAXe in his blog post [a big thing in a small package].
So, you have to wonder, did IBM XIV force EMC's hand into offering this new VMAXe storage unit? Surely, EMC sales reps will continue to lead with the more profitable DMX-4 or VMAX, and then only offer the VMAXe when the prospective customer mentions that the IBM XIV Gen3 is 69 percent less expensive. I haven't seen any list or street prices for the VMAXe yet, but I suspect it is less expensive than VMAX, on a dollar-per-GB basis, so that EMC will not have to discount it as much to compete against IBM.
technorati tags: IBM, XIV, Gen3, EMC, DMX-4, VMAX, V-Max, HDS, N7900, SONAS, DS8000, CKD, FICON, TCO
Well, it's Tuesday again, and that means IBM announcements! Right on the heels of our big storage launch on February 9, today IBM announced some exciting options for its modular disk systems. Let's take a look:
- 2TB SATA-II drives
That's right, you can now DOUBLE your capacity with 2TB SATA type-II drives on the DS3950, DS4200, DS4700, DS5020, DS5100 and DS5300 disk controllers, as well as the DS4000 EXP420, EXP520, EXP810, EXP5000 and EXP5060 expansion drawers. Here are the Announcement Letters for the [HVEC] and [AAS] ordering systems.
- 300GB Solid State Drives
IBM also announces 300GB solid state drives (SSD) for the DS5100 and DS5300. These are four times larger than the 73GB drives IBM offered last year, for those workloads that need high read IOPS such as Online Transaction Processing (OLTP) and Enterprise Resource Planning (ERP) applications. Here is the [Announcement Letter].
- New N series model N3400
For customers that need less than the minimum 21TB that our IBM Scale-Out Network Attach Storage (SONAS) can provide, IBM offers the new N3400 unified storage disk system, with support for NFS, CIFS, iSCSI and FCP. This is a 2U high 12 drive model that can be expanded up to 136 drives, basically doubling all the stats from last year's N3300 model. Fellow blogger, Rich Swain (IBM), does a great job recapping the speeds and feeds over on his blog [News and Information about IBM N series].
It also appears that the reports and rumors of the death of the DS6800 are premature. Don't believe misleading statements from competitors, such as those found written by fellow blogger BarryB (EMC), aka "the Storage Anarchist", in his latest post [Bring Out Your Dead] showing a cute little tombstone with "Feb 2010" on the bottom. Actually, if he had bothered to read IBM's [Announcement Letter], he would have realized that IBM plans to continue to sell these until June. Of course, IBM will continue to support both new and existing DS6800 customers for many years to come.
Technically, BarryB does not make any factually incorrect statements for me to correct on his blog. The idea that a product is "dead" is, of course, just opinion, and competitors poke fun at each others' announcements every day. One could argue that the EMC V-Max was "dead" after the ITG whitepaper [Cost/Benefit Case for IBM XIV Storage System - Comparing Costs for IBM XIV and EMC V-Max Systems] demonstrated that the IBM XIV cost 63 percent less than a comparable EMC V-Max over the life of three years total cost of ownership (TCO) back in July 2009. The comparison was made with data from clients in a variety of industries including manufacturing, health care, life sciences, telecommunications, financial services, and the public sector. This could explain why so many EMC customers are buying or investigating the IBM XIV and the rest of the IBM storage portfolio.
technorati tags: announcements
, Barry Burke
Am I dreaming? On his Storagezilla blog, fellow blogger Mark Twomey (EMC) brags about EMC's standard benchmark results, in his post titled [Love Life. Love CIFS.]. Here is my take:
- A Full 180 degree reversal
For the past several years, EMC bloggers have argued, both in comments on this blog, and on their own blogs, that standard benchmarks are useless and should not be used to influence purchase decisions. While we all agree that "your mileage may vary", I find standard benchmarks are useful as part of an overall approach in comparing and selecting which vendors to work with, and which architectures or solution approaches to adopt, and which products or services to deploy. I am glad to see that EMC has finally joined the rest of the planet on this. I find it funny this reversal sounds a lot like their reversal from "Tape is Dead" to "What? We never said tape was dead!"
- Impressive CIFS Results
The Standard Performance Evaluation Corporation (SPEC) has developed a series of NFS benchmarks, the latest, [SPECsfs2008] added support for CIFS. So, on the CIFS side, EMC's benchmarks compare favorably against previous CIFS tests from other vendors.
On the NFS side, however, EMC is still behind Avere, BlueArc, Exanet, and IBM/NetApp. For example, EMC's combination of Celerra gateways in front of V-Max disk systems resulted in 110,621 OPS with overall response time of 2.32 milliseconds. By comparison, the IBM N series N7900 (tested by NetApp under their own brand, FAS6080) was able to do 120,011 OPS with 1.95 msec response time.
Even though Sun invented the NFS protocol in the early 1980s, they take an EMC-like approach against standard benchmarks to measure it. Last year, fellow blogger Bryan Cantrill (Sun) gives his [Eulogy for a Benchmark]. I was going to make points about this, but fellow blogger Mike Eisler (NetApp) [already took care of it]. We can all learn from this. Companies that don't believe in standard benchmarks can either reverse course (as EMC has done), or continue their downhill decline until they are acquired by someone else.
(My condolences to those at Sun getting laid off. Those of you who hire on with IBM can get re-united with your former StorageTek buddies! Back then, StorageTek people left Sun in droves, knowing that Sun didn't understand the mainframe tape marketplace that StorageTek focused on. Likewise, many question how well Oracle will understand Sun's hardware business in servers and storage.)
- What's in a Protocol?
Both CIFS and NFS have been around for decades, and comparisons can sometimes sound like religious debates. Traditionally, CIFS was used to share files between Windows systems, and NFS for Linux and UNIX platforms. However, Windows can also handle NFS, while Linux and UNIX systems can use CIFS. If you are using a recent level of VMware, you can use either NFS or CIFS as an alternative to Fibre Channel SAN to store your external disk VMDK files.
- The Bigger Picture
There is a significant shift going on from traditional database repositories to unstructured file content. Today, as much as [80 percent of data is unstructured]. Shipments this year are expected to grow 60 percent for file-based storage, and only 15 percent for block-based storage. With the focus on private and public clouds, NAS solutions will be the battleground for 2010.
So, I am glad to see EMC starting to cite standard benchmarks. Hopefully, SPC-1 and SPC-2 benchmarks are forthcoming?
technorati tags: IBM, tape, EMC, Mark+Twomey, Storagezilla, CIFS, NFS, Celerra, V-Max, N7900, VMware, VMDK, Sun, Oracle, StorageTek, tape, benchmarks, SPEC, SPECsfs, SPECsfs2008, SPC, SPC-1, SPC-2, NetApp, FAS6080