Slashdot Mirror
No Hassle RAID 5 Implementations?
LambSpam asks: "I had a nightmare week (last week) with two of our servers running Intel's U3-1L RAID controller (RAID 5). Whenever there's a power outage in our building these controllers randomly mark one or more of the drives in the array offline (even with adequate UPS support), which means I have to manually mark them online and/or rebuild. Intel acknowledged the problem, but their solution involves updating the backplane's firmware, the controller firmware (destructive upgrade!), and even the firmware on our IBM drives in the array because they 'draw too much power' in certain conditions. I've only used one other RAID 5 implementation (MegaRAID), and it NEVER had these kinds of problems, whereas if you sneeze too hard around this U3-1L card it will go offline. Is this common with most hardware RAID implementations? What RAID 5 implementations works without hassle? What should I stay away from?"
I've never had any problems with the PERC (PowerEdge Raid Controller) in the Dells i (used) to use for Sendmail servers. That kind of limits your choices, though..
Xiotech, affiliated with Seagate, offers a Mag unite in fractional TB's, with highly customizable options. They're linked to servers with QLogic fiber-channel cards and are easy to setup; they even have knowledgable tech come on site for install.
I've had the same sort of problems with Adaptec AAA-133 RAID controllers, (which are in reality software based RAIDs with some HW support from the card,) and I've had the RAID management software for another major brand I do not recall the name of knock the RAID offline when it crashes, plus before long the hardware failed. The only hardware RAID controllers I've had success with, (and great success with at that,) are ICP Vortex controllers. My experiences with all of the above have with under Novell Netware, but at least from what is claimed on ICP Vortex's web site, they have great Linux support. And with the quality of their monitoring software, hardware, and even BIOS setup utility, I suspect they work perfectly under every operating system. They are also one of the few I've come across that only appear as one PCI device, so they can be used with normal motherboards with devices that do not support IRQ sharing.
I've also had performance problems with IBM drives in RAID 0/1, and especially RAID 5 setups. I contacted IBM tech support to see if any of the settings could be tweaked, but the response was the drives are not RAID optimized. I switched to Seagate drives, and subjectively I would say the performance quadrupled under heavy load.
I've used 2 Voyager 3100's with the fibre module.
Due to excessive server room heat, we did lose a drive, but data was fine. While it has Windows software to monitor it when connected via scsi, they didn't have anything for unix, so configs had to be done via telnet on its serial port.
Were I used to work (An all-windows shop) we used Adaptec RAID cards in all our "tower" based servers. Even the lower priced models (AAA-131U2) always performed without a hitch and we never had any problems with them at all. AMI's RAID controllers are real nice and all, but for the price it just wasn't worth it. The Adaptec solutions performed just as well and at a lower cost. You'd do good to check em out.
Now the 3200 RAID Controllers int he Compaq's, thats another diffrent story altogether.
We had roughly 2000 servers, operating 24/7 @ 67 degrees F. Two times a year we had a site shutdown. Every single time we had to bring everything back up we would have anywhere from 3-5 Compaq array controllers die. But never once did the low-buck Adaptecs crap out on us.
Looking for hardware (Currently need: Large Etch-a-Sketch) Have one? See my journal!
Dude, obviously you don't have adequate UPS support if the device does something when the power goes out. It is possible to get a better UPS that will flick over and back quick enough for all hardware not to notice.
--Giving to trolls for the benefit of us all
Just as a shot in the dark, I would suggest trying to upgrade the firmware on the drives first. At one of my old jobs, we used nothing but IBM drives, and we constantly had problems with the drives becoming marked as bad or off line, but simply pulling them and plugging them back in (hot swap) would bring them back. In our situation, we were using IBM Netfinity servers with IBM raid controllers. When we talked to IBM, they admitted there was a problem with the firmware on the drivers which would cause the drive to not spit out just one error whenever an event (even a simple read error) happened, but to spew them constantly, which made the raid controller mark the drive as bad. Seeing as it only takes a few minutes of downtime and is non-destructive, it might be worth a shot.
First, are you sure your UPS is a *TRUE* UPS? Even a lot of the 'high end' UPSes out there are still REALLY switched UPSes. This could very well be your problem.
The other one is something I've heard of (I'm not an electrical expert, but I'll try to explain). Larger (older installations, particularly) sites were wired for three-phase electricity. Over time, they split the phases for normal 110 volt usage. There is a chance where if the PC is connected to power on one phase, but the external unit is connected to power from a different phase, that the differential between the two can cause problems, due to the ground connection between the two through the cable shielding. I know, it sounds like something from the BOFH daily calendar, but it does make sense. Try making sure both pieces of equipment are on the same true UPS, or at least switched UPSes on the same circuit.
Unless you're limited by cost, don't use host based RAID. It will always be less reliable then a dedicated RAID controller. If you must use host based RAID, try and find a card that supports RAID 0/1 because it's faster and more reliable. I've had good experiences with MegaRAID cards, and the IBM host based raid controllers, but by good experience I mean that I've only had a few problems. There is always a chance that something will get screwed up when you change your setup.
Use a high end ICP Vortex controller with 15K RPM Cheetah SCSI drives or Fujitsu drives. Its the only combo I trust in any of my PC servers.
Alternatively you could try Sun's A3500 FCAL drive arrays with the 15K cheetahs for non PC hardware.
Sounds like good advice in the post above.
Some UPSs switch. Some are always online. You want the latter for a RAID array.
The second paragraph is important. Check your input power. Everything attached to your network should be wired to the same power circuit. Otherwise there is a possibility for feeding large spurious signals to your hardware through the power line.
Bush's education improvements were
When I took over my current job the last network team had overloaded the circuits in the server room. We've had 3 circuits trip and had servers drop hard. None of the Compaq SmartArray controllers had any problems recovering.
I suggest you also fix you power problem. The systems should have no idea power was lost to the building. If you are using a UPS and this is still happening, I'd find a better one.
I've never had trouble with Compaq and HP Raid-5's but Dell Raids randomly kill themselves (i.e. marking one drive as offline, and the whole raid dies... and Dell support doesn't know how the stuff works....)
I've used 3ware Escalade cards repeatedly, and never had any problems. I've only actually used RAID 5 once, but so far no worries with it. Of course, these are IDE RAID cards, which may not be acceptable if you have lots of SCSI drives already.
Randall.
Property law should use #'EQ, not #'EQUAL.
I have built serveral RAID configuration with IBM ServeRAID cotrollers. One RAID5 array (16 drives, 1 hot spare) that I've managed has had 2 drives fail in the past year; the only thing I've had to do is take the bad drive out, pop another one in and it is automatically marked as a hot spare.
I was expecting a hassle, but it was mind-blowing to see how easy it was. The cross-platform remote management utility is a plus too.
If you want bulletproof and are willing to pay for it, you wont go wrong with a sun A1000.
m l? cid=22455&parentId=67713
... we've had the 200+gb model running for almost 3 years now with no problems at all.
http://store.sun.com/catalog/doc/BrowsePage.jht
they range in size from 75gb to 436 gb, I work for an EDU so we get almost a 50% discount on them, but they are worth every penny
Religion is a gateway psychosis. -- Dave Foley
have you tried vinum or the linux raidtools?
i am running a vinum based r5 setup across 30 18G FC drives (i wget'd sandin's site as soon as i saw an article about his setup on bp6.com) without any problems.
then again, power going out where i work (a power plant) is rather unlikely =)
Where I work all we use is Mylex cards for both 4 and 6 drive raid-5 implementations. We use IBM drives, but because of bad experiences lately (6 of them blowing up) we've recently switched to Seagate Cheetahs U have to drop your kernel down (we run stably under 2.2.12, and have had problems getting it to work on 2.2.18) but if you're running on linux-based servers, mylex is the way to go You can get both 32 and 64 bit PCI cards, and at only about 3-4 grand CDN a pop... it isn't that costly for a hardware RAID-5
Hitachi or EMC storage array. Built in UPS that will flush the cache to disk as a last resort before power can fail. And hitachi tells me they'll do linux. Awesome performance too. Money? If you have to ask....
--Chris
We had this exact problem on our servers at work and it was a real headache getting them upgraded to the new firmware. It's a serious problem and it's imperative that you upgrade to the newest rev. of the firmware, not just the patch.
e rv er/ta_445.htm
Intel's site has a technical advisory dated Jan 29th, 2002 regarding drives being 'marked offline".
http://support.intel.com/support/motherboards/s
All 110v circuits are split off of two or three phase.
Where are you from, some place where the generate using some type of static electrical machine?
It's supposed to be fast, cheap, and reliable....
The Dell PERC (actually, I forget the company who actucally makes them as Dell just masks them as Dell) logic is written primarily by Intel. My co-worker's husband wrote a lot of code for the 3/ series. I'll be sure to tell her to give her husband some shit. =)
These are under $300 on Ebay, work great, and have many features. You'll have better compatibility experience with AMI cards, Mylex have more features though, but older eXtremeRAID have proprietary memory modules (which will cost $1000 retail if you want upgrade, if you find one somewhere).
ICP Vortex have great reputation, though I don't have any experience with them.
Everything needs to be on the same Ground circuit. It is necessary to avoid ground loops.
"They draw LARGE spikes of current sporadically."
I don't think this is correct. I have designed power supplies, and I don't immediately think of any reason why the power input of a switching power supply should vary differently from the power output. The only surge is when the hard disks spin up, but with SCSI there is a means to stagger the spin-up.
Bush's education improvements were
In this situation, I use XML. I invent my own markup language that is self-consistent and describes the API of a system. I then use an XSLT processor, Apache Xalan to be precise, to transform the source to various other formats including: a web site, one big printable web page, PDF, and I've been thinking about writing a stylesheet for man pages as well.
The only issue with a system like this is version control of your source files, which is highly situation specific.
std::disclaimer<std::legalese> sig=new std::disclaimer; sig->dump(); delete sig;
We have used Raidtec boxes for quite a long time, and they have always been very reliable.
I think all of our Raidtecs are kitted out with Seagate drives.. anyway, check out http://www.raidtec.com for a little more information on what they sell.
"Hey! Unless this is a nude love-in, get the hell off my property!!"
... is Compaq.
I've probably set up over 100 servers over the last 10 years or so, and I wouldn't use anything but Compaq Array controllers. I've never lost data because of a drive subsystem problem. I've got over 20 that I'm responsible for now, and all of them use Compaq Array controllers. They are reliable, easy to configure, well supported, and easy to maintain. The tools under NetWare and Windows work well. Most are supported under Linux. They aren't cheap,but they are simply great.
For details look here.
I have worked for one large regional financial institution, and one large entertainment conglomerate, and one of the things they have in common is that both use Compaq hardware. There's a good reason - it works.
FWIW, I do not now, nor have I ever worked for Compaq, nor do I have any direct investment in Compaq.
-- "Never underestimate the power of human stupidity." - R.A.H.