Обсуждение: Perc 3 DC
Hi chaps, I've had this old card sitting on my desk for a while. It appears to be a U160 card with 128Mb BBU so I thought I'd wangit in my test machine (denian etch) and give it a bash. I set up 4 36Gb drives in raid 0+1, but I don't seem to be able to get more than 20MB/s write speed out of it for large files(2XRAM usual tests with dd from dev/zero). I don't expect anything great, but I thought it'd do a little better thanthat. I've tried writeback and write through modes, tried changing cache flush times, disabled and enabled multiple PCI delayedtransactions, all seem to have little effect. Finally I decided to wave goodbye to Dell's firmware. LSI has it down as a MegaRAID 493 elite 1600, so I flashed it withtheir latest firmware. Doesn't seem to have helped either though. Has anybody else used this card in the past? I'm wondering if this is a driver issue, or if the card is and always was justcrap? If so I'll proabably try sw raid on it instead. Any thoughts? Glyn
On Sat, Nov 22, 2008 at 7:18 AM, Glyn Astill <glynastill@yahoo.co.uk> wrote: > Hi chaps, > > I've had this old card sitting on my desk for a while. It appears to be a U160 card with 128Mb BBU so I thought I'd wangit in my test machine (denian etch) and give it a bash. > > I set up 4 36Gb drives in raid 0+1, but I don't seem to be able to get more than 20MB/s write speed out of it for largefiles (2XRAM usual tests with dd from dev/zero). I don't expect anything great, but I thought it'd do a little betterthan that. You really have two choices. First is to try and use it as a plain SCSI card, maybe with caching turned on, and do the raid in software. Second is to cut it into pieces and make jewelry out of it. Anything before the Perc 6 series is seriously brain damaged, and the Perc6 brings the dell raid array line squarly in line with a 5 year old LSI megaraid, give or take. And that's being generous. > I've tried writeback and write through modes, tried changing cache flush times, disabled and enabled multiple PCI delayedtransactions, all seem to have little effect. Yeah, it's like trying to performance tune a yugo. > Finally I decided to wave goodbye to Dell's firmware. LSI has it down as a MegaRAID 493 elite 1600, so I flashed it withtheir latest firmware. Doesn't seem to have helped either though. Does it have a battery backup module? Often you can't really turn on write-back without one. That would certainly slow things down. But you should certainly expect > 20 M/s on a modern RAID controller writing out to a 4 disk RAID10 > Has anybody else used this card in the past? I'm wondering if this is a driver issue, or if the card is and always wasjust crap? If so I'll proabably try sw raid on it instead. It's pretty much a low end card. Look into modern LSI, Areca, or Escalade/3Ware cards if you want good performance.
--- On Sat, 22/11/08, Scott Marlowe <scott.marlowe@gmail.com> wrote: > You really have two choices. First is to try and use it as > a plain > SCSI card, maybe with caching turned on, and do the raid in > software. > Second is to cut it into pieces and make jewelry out of it. Haha, I'm not really into jewelry, although I had thought of smacking it into a pile of dust with a lump hammer, that's muchmore my thing. > Anything > before the Perc 6 series is seriously brain damaged, and > the Perc6 > brings the dell raid array line squarly in line with a 5 > year old LSI > megaraid, give or take. And that's being generous. > Well this card thinks it's a 5 year old lsi megaraid. I've got a pile of perc5i megaraid paperweights on my desk at work,so this was kinda expected really. > > I've tried writeback and write through modes, > tried changing cache flush times, disabled and enabled > multiple PCI delayed transactions, all seem to have little > effect. > > Yeah, it's like trying to performance tune a yugo. > Did I mention I drive a yugo? > > Finally I decided to wave goodbye to Dell's > firmware. LSI has it down as a MegaRAID 493 elite 1600, so I > flashed it with their latest firmware. Doesn't seem to > have helped either though. > > Does it have a battery backup module? Often you can't > really turn on > write-back without one. That would certainly slow things > down. But > you should certainly expect > 20 M/s on a modern RAID > controller > writing out to a 4 disk RAID10 > Yeah the battery's on it, that and the 128Mb is really the only reason I thought I'd give it a whirl.
On Sat, Nov 22, 2008 at 9:59 AM, Glyn Astill <glynastill@yahoo.co.uk> wrote: > --- On Sat, 22/11/08, Scott Marlowe <scott.marlowe@gmail.com> wrote: > >> You really have two choices. First is to try and use it as >> a plain >> SCSI card, maybe with caching turned on, and do the raid in >> software. >> Second is to cut it into pieces and make jewelry out of it. > > Haha, I'm not really into jewelry, although I had thought of smacking it into a pile of dust with a lump hammer, that'smuch more my thing. Well, I think the important thing here is to follow your bliss. :) >> Anything >> before the Perc 6 series is seriously brain damaged, and >> the Perc6 >> brings the dell raid array line squarly in line with a 5 >> year old LSI >> megaraid, give or take. And that's being generous. >> > > Well this card thinks it's a 5 year old lsi megaraid. I've got a pile of perc5i megaraid paperweights on my desk at work,so this was kinda expected really. Yeah we just wound up buying a RAID controller from Dell for handling 8 500Gig drives and no one checked and they sold us a Perc5, which can only handle a 2TB array apparently. Set it up as a RAID-10 and it's definitely got "meh" levels of performance. I had an old workstation with a 4 port SATA card (no raid) running software raid and it handily stomps this 8 disk machine into the ground. >> > I've tried writeback and write through modes, >> tried changing cache flush times, disabled and enabled >> multiple PCI delayed transactions, all seem to have little >> effect. >> >> Yeah, it's like trying to performance tune a yugo. > > Did I mention I drive a yugo? Well, they're fine vehicles for what they do (I hear they're still quite an icon in eastern europe) but they aren't gonna win a lot of cash at the race track. :) >> down. But >> you should certainly expect > 20 M/s on a modern RAID >> controller >> writing out to a 4 disk RAID10 >> > > Yeah the battery's on it, that and the 128Mb is really the only reason I thought I'd give it a whirl. We had a bunch of 18xx series servers last company I was at (we went from unix / linux to Microsoft, so ordered some 400 machines to replace a dozen or so unix machines) and they all came with the adaptec based Perc 3s. We wound up with a large cardboard box full of them and running software RAID to get decent performance and to stop them from locking up randomly. I had the only two running linux and equipped with LSI based Perc3. They were stable, but the pair of intel OEM boxes they replaced, which had 1/4 the CPU horsepower, were still noticeably faster with their older but LSI based RAID controllers. I left that place as soon as I could, I just couldn't handle a large portion of my job being to walk around a data center resetting locked up windows boxes each morning. I think as much as anything the busses on the dells are the problem, resulting in pretty poor throughput, especially true of the old serverworks chipset machines. Those things are pretty much boat anchors.
--- On Sat, 22/11/08, Scott Marlowe <scott.marlowe@gmail.com> wrote: > I had an old workstation with a 4 port SATA card (no raid) running > software raid and it handily stomps this 8 disk machine into the ground. Yeah, I think this machine will be going that route. > We had a bunch of 18xx series servers last company I was at > (we went > from unix / linux to Microsoft, so ordered some 400 > machines to > replace a dozen or so unix machines) I'm not surprised. We've just had some management "inserted" to make decisions like that for us. Honestly if I get askedone more time why we're not utilizing iSCSI or <insert buzzword here> more .... But that's another matter. > > I think as much as anything the busses on the dells are the > problem, > resulting in pretty poor throughput, especially true of the > old > serverworks chipset machines. Those things are pretty much > boat > anchors. Funny that, possibly explains some of the useless supermicro hardware I had a while back.
Glyn Astill wrote: > --- On Sat, 22/11/08, Scott Marlowe <scott.marlowe@gmail.com> wrote: > > >>You really have two choices. First is to try and use it as >>a plain >>SCSI card, maybe with caching turned on, and do the raid in >>software. >>Second is to cut it into pieces and make jewelry out of it. > > > Haha, I'm not really into jewelry, although I had thought of smacking it into a pile of dust with a lump hammer, that'smuch more my thing. > > >> Anything >>before the Perc 6 series is seriously brain damaged, and >>the Perc6 >>brings the dell raid array line squarly in line with a 5 >>year old LSI >>megaraid, give or take. And that's being generous. >> > > > Well this card thinks it's a 5 year old lsi megaraid. I've got a pile of perc5i megaraid paperweights on my desk at work,so this was kinda expected really. > > >>>I've tried writeback and write through modes, >> >>tried changing cache flush times, disabled and enabled >>multiple PCI delayed transactions, all seem to have little >>effect. >> >>Yeah, it's like trying to performance tune a yugo. >> > > > Did I mention I drive a yugo? > > >>>Finally I decided to wave goodbye to Dell's >> >>firmware. LSI has it down as a MegaRAID 493 elite 1600, so I >>flashed it with their latest firmware. Doesn't seem to >>have helped either though. >> >>Does it have a battery backup module? Often you can't >>really turn on >>write-back without one. That would certainly slow things >>down. But >>you should certainly expect > 20 M/s on a modern RAID >>controller >>writing out to a 4 disk RAID10 >> > > > Yeah the battery's on it, that and the 128Mb is really the only reason I thought I'd give it a whirl. > > > > Is the battery functioning? We found that the unit had to be on and charged before write back caching would work.
--- On Mon, 24/11/08, Steve Clark <sclark@netwolves.com> wrote: > > Yeah the battery's on it, that and the 128Mb is > really the only reason I thought I'd give it a whirl. > > > > > Is the battery functioning? We found that the unit had to > be on and charged before write back caching > would work. Yeah the battery is on there, and in the BIOS it says it's "PRESENT" and the status is "GOOD".
On Mon, Nov 24, 2008 at 7:49 AM, Glyn Astill <glynastill@yahoo.co.uk> wrote: > --- On Mon, 24/11/08, Steve Clark <sclark@netwolves.com> wrote: > >> > Yeah the battery's on it, that and the 128Mb is >> really the only reason I thought I'd give it a whirl. >> > >> > >> Is the battery functioning? We found that the unit had to >> be on and charged before write back caching >> would work. > > Yeah the battery is on there, and in the BIOS it says it's "PRESENT" and the status is "GOOD". If I remember correctly, older LSI cards had pretty poor performance in RAID 1+0 (or any layered RAID really). Have you tried setting up RAID-1 pairs on the card and then striping them with the OS?
--- Scott Marlowe <scott.marlowe@gmail.com> wrote: > > > > Yeah the battery is on there, and in the BIOS it says it's > "PRESENT" and the status is "GOOD". > > If I remember correctly, older LSI cards had pretty poor > performance > in RAID 1+0 (or any layered RAID really). Have you tried setting > up > RAID-1 pairs on the card and then striping them with the OS? > Not yet no, but that's a good suggestion and I do intend to give it a whirl. I get about 27MB/s from raid 1 (10 is about the same) so hopefully I can up the throughput to the speed of about one disk with sw raid. For kicks I did try raid 5 on it; 6.9MB/s made it hard to resist going to get the hammer, which is still a very attractive option.
On Mon, Nov 24, 2008 at 8:06 AM, Glyn Astill <glynastill@yahoo.co.uk> wrote: > --- Scott Marlowe <scott.marlowe@gmail.com> wrote: > >> > >> > Yeah the battery is on there, and in the BIOS it says it's >> "PRESENT" and the status is "GOOD". >> >> If I remember correctly, older LSI cards had pretty poor >> performance >> in RAID 1+0 (or any layered RAID really). Have you tried setting >> up >> RAID-1 pairs on the card and then striping them with the OS? >> > > Not yet no, but that's a good suggestion and I do intend to give it a > whirl. I get about 27MB/s from raid 1 (10 is about the same) so > hopefully I can up the throughput to the speed of about one disk with > sw raid. > > For kicks I did try raid 5 on it; 6.9MB/s made it hard to resist > going to get the hammer, which is still a very attractive option. Well, I prefer making keychain fobs still, but from a technical perspective, I guess either option is a good one. Srsly, also look at running pure sw RAID on it with the controller providing caching only. I don't expect a PERC 3DC to win any awards, but the less you give that card to do the better off you'll be.
On Monday 24 November 2008 14:49:17 Glyn Astill wrote: > --- On Mon, 24/11/08, Steve Clark <sclark@netwolves.com> wrote: > > > Yeah the battery's on it, that and the 128Mb is > > > > really the only reason I thought I'd give it a whirl. > > > > > > Is the battery functioning? We found that the unit had to > > be on and charged before write back caching > > would work. > > Yeah the battery is on there, and in the BIOS it says it's "PRESENT" and > the status is "GOOD". Sorry I deleted the beginning of this on getting back from a week off. Writeback is configurable. You can enabled write back caching when the unit is not charged if you like. It is offered when you create the array (and can be changed later). It is arguably a silly thing to do, but it is an option. I have some reasonable performance stats for this card assuming you have a suitably recent version of the driver software, DELL use to ship with a Linux kernel that had a broken driver for this card resulting is very poor performance (i.e. substantially slower than software RAID). I have a note never to use with Linux before 2.6.22 as the LSI driver bundled had issues, DELL themselves shipped (if you asked "why is performance so bad") a Redhat kernel with a later driver for the card than the official Linux kernel. That said a couple of weeks back ours corrupted a volume on replacing a dead hard disk, so I'm never touching these cheap and tacky LSI RAID cards ever again. It is suppose to just start rebuilding the array when you insert the replacement drive, if it doesn't "just work" schedule some down time and figure out exactly why, don't (for example) blindly follow the instructions in the manual on what to do if it doesn't "just work".
On Mon, Nov 24, 2008 at 8:41 AM, Simon Waters <simonw@zynet.net> wrote: > That said a couple of weeks back ours corrupted a volume on replacing a dead > hard disk, so I'm never touching these cheap and tacky LSI RAID cards ever > again. It is suppose to just start rebuilding the array when you insert the > replacement drive, if it doesn't "just work" schedule some down time and > figure out exactly why, don't (for example) blindly follow the instructions > in the manual on what to do if it doesn't "just work". Reminds me of a horror story at a company I was at some years ago. Another group was running Oracle on a nice little 4 way Xeon with a Gig of ram (back when they was a monster server) and had an LSI card. They unplugged the server to move it into the hosting center, and in the move, the scsi cable came loose. When the machine came up, the LSI RAID marked every drive bad and the old 4xx series card had no facility for forcing it to take back a drive. All their work on the db was gone, newest backup was months old. I'm pretty sure they now understand why RAID5 is no replacement for a good backup plan. I had a 438 in a dual ppro200, and it worked just fine, but I never trusted it to auto rebuild anything, and made backups every night. It was slow (in the 30 meg/second reads on a 6 disk RAID 5, not faster in RAID-10 for reads or writes) but reliable. Newer LSI cards seem quite nice, but I'm now using an Areca 16xx series and am so far very happy with it's reliability and somewhat happy with its performance. Sequential read speed is meh, but random performance is very good, so for a db server, it's a nice unit.
> Not yet no, but that's a good suggestion and I do intend to give it a
> whirl. I get about 27MB/s from raid 1 (10 is about the same) so
> hopefully I can up the throughput to the speed of about one disk with
> sw raid.
FYI I get more than 200 MB/s out of a Linux Software RAID5 of 3 SATA
drives (the new Samsung Spinpoints...)
(Intel ICH8 chipset, Core 2 Duo).