"two triple disk RAID 6 failures"
I'd like to learn more about this. And also how ISE achieve failure rate of 0.1%
X-IO to heat up ISE storage bricks with iSCSI access
Sealed disk/flash array supplier X-IO is growing at 30 per cent a year and intending to widen its market by adding iSCSI later this year to the existing Fibre Channel access. ISE and Hyper ISE products are 3U sealed rack enclosures requiring no maintenance access for five years as the smart software inside copes with any …
-
-
Thursday 13th March 2014 12:11 GMT Steven Jones
Re: "two triple disk RAID 6 failures"
It's fairly easy to see how they achieve such low failure rates. They simply put the redundancy in the package and put in a system than supports multiple disk failures, probably via some sort of virtual RAID. That way you could provide full access, even if two of the actual drives have suffered physical failures.
Of course, you can't replace the internal drives yourself to restore the full resilience - it's a sealed package. After 5 years, you could well start suffering more failures if some of the internal redundancy has been compromised.
The secret here is the failure rate is measured by the failure of the entire ensemble failing, not the rate of failure of internal disks. The proper comparison is that of the failure rate of a RAID array, and not an individual disk.
-
Thursday 13th March 2014 12:15 GMT Nate Amsden
Re: "two triple disk RAID 6 failures"
last I spoke to them 5 yrs ago they did it by re certifying the drives in the array. since most failures are not real failures. for example most vendors fail a drive forcefully after some sumber of read errors. Xio had/has seagates special rma recertification tech built right in(nobody else does).
if a drive has a lot of failed bits then xio can fail individual platters and keep the rest of the drive going. There is enough spare capacity in the system that overall usable capacity is not affected. el reg had a good article on Xio's tech a few years ago(maybe 6).
at the time they claimed their field data showed they could offer 7 year warranties but the decided to go for 5 instead.
I have never used their stuff but it seems they have always been hampered by lack of features and poor software relative to say the traditional tier 1 folks.
the tech behind their disks is pretty cool but not enough to get someone to acquire them or significantly grow. They missed the boat.
I have been very very loosely following them for many years. Still a die hard 3par customer though.
(blame errors on my phone kb)
-
Thursday 13th March 2014 14:04 GMT Steven Jones
Re: "two triple disk RAID 6 failures"
Their website describes it quite well, although they do use some dubious terminology. For instance, under self-healing they claim the disk is "remanufactured"
"using our patented software technology, every drive in an ISE can be taken into repair mode, power-cycled, and remanufactured just as it would be if it were RMA’d to the drive’s supplier — all while the ISE continues to operate at full performance and full capacity."
That's a weird use of the term "remanufactured". Of course what they will really be doing is performing a full low-leve format, identify bad tracks/sectors and so on. It's hardly the normal use of the word "remanufactured", which usually means replacing parts of the device (like bearings) which have fallen out of manufacturing tolerances. Of course, the can get away with these weasel terms as no manufacture actually remanufactures disk drives these days - they simply aren't seviceable that way. So yes, they might do what a manufacturer does when faced with a returned disk, but remanufacturing it isn't. I think re-certifying, as you had it described to you, is far more accurate.
However, they also say that if part of the disk is unusable then they will take that element out of service. I've no doubt that extends to a full HDD failing when the entire disk will have to be retired. At that point there's no way a user can restore it to full resilience as the units are sealed.
nb. the likes of NetApp have schemes which monitor drives and dispatch replacements to be user-replaced (not a hard job). The returned drives will then be assessed and go through reformats, map-outs and re-certification. If they don't meet standards, they can be taken out of the pool of replacements for dispatch to customers. I've no idea of the exact methods used, but I suspect it's possible to do more extensive work back at the factory so to speak. I doubt what NetApp do is much different in principle to other enterprise storage suppliers.
-
-
-
Thursday 13th March 2014 22:37 GMT X-IO Hollis
High Reliability… It’s not that easy…
Let me say that I welcome the comments, and appreciate the opportunity to talk about how the ISE can achieve these high levels of reliability. If we were simply another “storage” company that uses the cheapest commodity components we could find, there is obviously only so much that can be done with individual disk drives. This approach assumes that HDDs (and SSDs for that matter) are simply disposable components that are inherently unreliable. Unfortunately it’s this very approach that causes the high failure rates that we see in enterprise environments. By putting drives in commodity chassis with plastic casings (LOTS of vibration/horrible cooling), and simply treating them as “holes” that the controllers throw data at, the overall system is made much less reliable.
Check out the below spec sheet for a Seagate enterprise HDD:
http://www.seagate.com/support/internal-hard-drives/enterprise-hard-drives/savvio-10k/
On the 2nd page in the chart, Seagate states that these drives have an Annual Failure Rate of 0.55%, but when storage companies put them in commodity enclosures that failure rate goes to 4% - 7%. Why? They would claim their software actually improves the reliability, but how is this better?
We are not that…
The way you increase reliability is to fix the things that are causing the HDDs to “fail” in the first place. This is not an easy thing to do, and it took the storage team from Digital Equipment Corporation (who was then Compaq, who was then HP) and the engineers from Seagate (yup… That Seagate) to redesign how storage arrays are designed. From custom designing the ideal physical environment for the HDDs, changing the way that drives operate in groups, and being able to fail sub-components of the drive (individual Heads or parts of a platter), the ISE represents the best of what can be done with storage when the leading creators of the technology come together. X-IO has many customers that have been in operation for 5 years without a Service Event:
http://xiostorage.com/x-io-technologies-unique-5-year-zero-touch-storage-proven/
Please reach out to us at fastforever@x-io.com if you would like more information. We would be happy to talk about how the ISE can increase the reliability of your environment.
Gavin McLaughlin (interviewed for this article), has an excellent post on reliability and HDDs
http://xiostorage.com/taint-what-you-do-its-the-way-that-you-do-it/
This topic is closed for new posts.
Biting the hand that feeds IT
