Camera Labs

Remember when computers worked?

haha!

I've had no issues with my i7 rig but another I've built is giving greif Constantly resetting we think its either the PSU or the Motherboard, although we're 80% sure its the PSU, changing it next week.

Also Bob, I hate using USB hubs! They confuse things >_< I'd prefer to just use ones on the front/top/side of the case or just use longer cables for things.

Also are you planning on using the SAS controller for SAS drives? Or just for SATA ones? And also im assuming the storage controllers on the P6T Deluxe and WS PRO are the same - is it possible to set up independant raids on the SAS and SATA controllers?

AND

Is it possible to set up more than 1 raid on the SATA controller, e.g.

SATA 6 Ports - 2 Drives - Raid 1, 2 Drives - Raid 0, 2 Drives - JBOD

SAS 2 Ports - 2 Drives - JBOD

_________________
Canon EOS 5DmkII + BG-E6 + Canon EOS 40D + BG-E2N + Canon EOS 33
Canon 17-40 F/4L USM + Canon 24-70 F/2.8L USM + Canon 28mm F/1.8 USM + Canon 70-200 F/2.8L USM IS + Canon 85mm F/1.8 USM
Canon Speedlite 580EX II + Canon Speedlite 540EZ + 2 x Nikon SB-80DX
Cactus V2s Wireless Trigger - 5 x Cactus V2s Wireless Reciever

MY FLICKR!

Hi Daniel,

It is possible to set up RAID on both the onboard and SAS controllers. The tricky part is to get RAID working on the SATA side while enabling the SAS side to control the boot disk, as detailed in my earlier post. I decided to dedicate the SAS controller to my single SATA SSD as I thought the potentially high bandwidth requirement of that device better split from my SATA RAID 10 array. That left just enough onboard SATA ports to run the four HDD as RAID 10, an optical drive and the eSATA connector so I didn't need to invest in a separate SATA controller card.

So, AFAIK, it should be quite possible to set up your disks as you describe.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Hi folks,

A cautionary tale about flashing the BIOS follows:

As detailed above I have my Asus P6T Deluxe motherboard configured slightly unusually with an SSD SATA drive hanging off the Marvell SAS controller. This drive is the system's C: drive and holds the OS. Four SATA drives configured as a RAID array provide mass storage.

So far so good but I thought I'd update the BIOS today in preparation for Windows 7. Didn't really have to, of course, but I'd missed the last few revisions which did include a few bug fixes. The update went wonderfully, I rebooted and the Marvell SAS controller ROM, when AHCI was re-enabled as required to access my SSD, was hanging the boot sequence. Shouldn't be a problem except that the hang was occurring before getting into the BIOS setup so it was impossible to troubleshoot from the keyboard.

I must have had the machine out and disassembled close on a dozen times today, a procedure made all the worse because the CMOS reset jumper was underneath the graphics card so that had to come out every time as well.

Still no joy and then I discovered a helpful hint on the Asus BIOS download page saying "Please update Marvell 88SE6320 SAS Controller Firmware update Version 3.1.0.2300 after Bios update" (my bold text). Huh? Chance would be a fine thing! Then I found an FAQ entitled "My system can’t pass POST if I enable AHCI function in Bios setting after updated Bios". An FAQ, mind you, on the Asus web-site!!!

In the end I had to create a DOS bootable USB drive and, after a final CMOS reset, reflash to the latest BIOS (I'd reverted to the previous version but the problem persisted) and then update the Marvell ROM from a special utility downloaded from the Asus site. Doh!

Well, they do say that if it ain't broke don't fix it but I do take exception at Asus' rather cavalier attitude. Either issue a BIOS update which doesn't break something else (the SAS controller) or add a caution to the BIOS download page inviting one to perform an SAS firmware update before updating the main BIOS.

Hope this helps anyone else with this (Asus P6T Deluxe) motherboard (or similar?) from wasting their day.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

That happened to me...

the other month when i got those 2 nice Samsung F3's and installed windows 7 I did some re-arranging or hard drives etc.

I thought ah what the hey, windows 7 ill lob the AHCI on. haha. Then I had the problem with the not posting or anything so had to reset the CMOS, luckily I figured this out to be the case within minutes and my reset switch on my P6TWS is easy to get too.

I'm not fussed about the AHCI becasue I don't really need it and there isnt any real speed benifit.. If I get a RAID5/6 at any point then ill probably be buying a dedicated hardware RAID card from 3Ware or the like, I'm not a huge fan of the intel controller for anything but just single drives.

I never did check for a fix but it'd be a safe bet since its the same SAS controller it'll most likely be the same if not a similar update to the controller. I'll keep it in mind next time I've got the system up on my bench for its next clean out.

_________________
Canon EOS 5DmkII + BG-E6 + Canon EOS 40D + BG-E2N + Canon EOS 33
Canon 17-40 F/4L USM + Canon 24-70 F/2.8L USM + Canon 28mm F/1.8 USM + Canon 70-200 F/2.8L USM IS + Canon 85mm F/1.8 USM
Canon Speedlite 580EX II + Canon Speedlite 540EZ + 2 x Nikon SB-80DX
Cactus V2s Wireless Trigger - 5 x Cactus V2s Wireless Reciever

MY FLICKR!

AHCI does add the benefit of native command queuing, something that can speed up data read and write speeds. Apart from that there are no real benefits, but as things are, if you're using an Intel chipset motherboard, you can't install the ACHI drivers on their own after you've installed the OS, which means you can't enable ACHI at a later stage. On an AMD or Nvidia based motherboard, you can install this once the OS have been installed and this allows you to enable ACHI at a later stage.
You're not having the same issues as with the SAS controller, it's just a matter of driver support and if you install the drivers at the same time as you install the OS, it'll work just fine.

No no you misunderstand, the system wont finish posting, it doesnt even get as far as trying to boot any OS. regardless if AHCI was on or not to start with, if its on and the SAS controller is on too and your trying to boot from it, be it a SAS or SATA drive, it wont have it.

Though me having a Dell Perc5/i SAS raid card in there too wont help clear up matters haha

_________________
Canon EOS 5DmkII + BG-E6 + Canon EOS 40D + BG-E2N + Canon EOS 33
Canon 17-40 F/4L USM + Canon 24-70 F/2.8L USM + Canon 28mm F/1.8 USM + Canon 70-200 F/2.8L USM IS + Canon 85mm F/1.8 USM
Canon Speedlite 580EX II + Canon Speedlite 540EZ + 2 x Nikon SB-80DX
Cactus V2s Wireless Trigger - 5 x Cactus V2s Wireless Reciever

MY FLICKR!

Hi Daniel,

Sorry you got caught as well. But the Marvell firmware upgrade procedure I described above works a treat and if, unlike me, you don't have to create a "DOS" bootable memory stick the fix only takes a few minutes and then you can get the Asus board configuration exactly how it was before Asus broke it.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

.
Hmm. I'll share this in the hope that nobody else gets caught out. What follows is both technical and very system specific so you may want to look away now...

As detailed above I have four SATA drives configured as RAID 10. My setup is slightly non-standard in that in order to get my SSD as the bootable drive when it is connected to the SAS controller I have to reconfigure the disk sub-system as AHCI (rather than RAID) from the main BIOS once the RAID array is created. It's a kludge but the workaround has been well documented elsewhere on the Net.

The reason I'm mentioning this once more is that I had a catastrophic failure of that RAID 10 array after which Windows 7 would reboot and fail to see the D: drive. Going into the motherboard BIOS I reconfigured the disk subsystem as RAID which allowed me to gain access to the ICH10R setup during boot-up by pressing Ctrl+I. That showed me that two disks had "an error". It also offered to fix one of them but not the other. The fix, whatever it was, was instant and allowed me, after setting up AHCI once more in the main BIOS and tweaking the boot order, to boot once more into Windows. During the boot CHKDSK started automatically and gave my D: drive a thorough mauling. Net result was that I lost a lot of files from that drive. Also, running the Intel Storage Manager console from Windows still shows that one of my four drives has an error.

While able to recover my data (I hope) from the network attached storage there are a number of programs which aren't so happy and need reinstallation. At that point I thought it better to use a System Image created by the Windows 7 backup utility last Sunday but, after the requested reboot, when the utility ran it very soon popped up an internal error so that was a bust.

So what have I learnt? First and foremost is that neither the BIOS nor the Intel software have the ability to alert me to a disk failure unless I go looking. The BIOS config is non-standard so I'll excuse that one but it seems bonkers that the Intel driver which manages the array can't provide a system alert. Second, and just as bad really, the System Image backup facility provide with Windows, now in it's second iteration with W7, is worse than useless in that it suckers one into believing that you are backed up but give it anything a bit more complicated than a simple drive setup and it fails to do what it says on the tin. I've used it with success on a PC with a single SATA drive but there's no way I have any confidence in it for the main machine. Thank goodness I had another backup of my data on the NAS using a third party file synchroniser. Sometimes paranoia pays off.

Hopefully this story will save someone else from disaster. It probably applies to very few but if it helps just one person it will have been worth setting down.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

I run "only" raid 1 in a conventional setup. The Intel utility does certainly work in that case, and flashes up dire warnings if for example I forget to plug one of the disks in during tinkering. Similarly if I have a bad system crash (only happened once), where I guess write data may not be in sync, it marks the mirror as being out of sync and rebuilds automatically on next boot.

Out of interest, what was the cause of the failure in the first place? Is one (or more?) of the disks really bad, or did it just go wobbly for some unknown reason?

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Hi popo,

I'm still pretty much in the dark. What I did see occasionally over the preceding week was an unsolicited CHDSK offer on the D: drive just before the W7 Log On screen at boot up. With hindsight that may have been due to a first disk failure in the array but CHDSK never found any problem to correct and when I examined the drive using the Windows "Disk Management" tool it always flagged as healthy. Maybe if I had run Intel's Storage Management console I'd have seen a different story.

This morning the crash was signalled by an inability to perform some Windows operations. Firefox froze on me and wouldn't shut down. Requesting Windows to force a shut down of Firefox didn't help either. In the end I told Windows itself to restart and it hung for some time at the Shutting Down screen as well. In fact I did the patient thing and went to make a cup of tea and at some stage during the tea making process the Log On screen duly appeared. It was only after logging on that I discovered the lack of a D: drive.

As mentioned above, once I got to the correct BIOS page for the Intel controller it was quickly able to "fix" one of the drives but the other still remains flagged with "Unknown Status" (by ISM). My working guess is that with one drive having "silently" failed" just a momentary glitch on the adjacent drive was enough to lose the entire RAID 10 D: drive and as that disk holds the paging file and a lot of the User AppData files normally found on a C: drive (I offloaded them to save wear and tear on the SSD) that was enough to bring Windows down in a major way.

A replacement drive should be delivered tomorrow and then I'll have the fun (!) of trying to figure out which one of the four drives needs to be pulled. Hopefully the motherboard SATA port labelling will be clear or it will be trial and error. Then I'm hoping that the ICH10R firmware will offer to rebuild the array from the three good drives. Optimistic, maybe, but I don't see any way of telling it to do this from a menu.

But one thing for sure is that CHKDSK did give my D: drive a thorough mauling after the first failed drive was restored. While most of the files appear to be intact and the right size some are actually gibberish and I've had to delete them and copy good versions across from the network storage.

For the medium term even if I do get the computer fully healthy again I may relegate the on-board SATA controller to optical duties and invest in a controller card. This machine is the first one I've built in a long time to use the on-board solutions and I'm really regretting that decision now even though it was based on an eco-friendly desire to minimise power consumption. At least if I do go for a separate card it will give me the opportunity to install W7 as a fresh install rather than the upgrade over Vista which I currently have but I'm going to have to plan that one very carefully.

Bob.

P.S. Firefox froze while I was reading Gordon's 550D/T2i preview. It's all his fault.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Assuming it behaves like a mirror, in ISM it'll show you which disks are in the array and any spare ones you can assign to the array for repair. Hope it goes smoothly as can be.

On the crash timing, maybe the controller was reading the review with you and decided 18MP was too much for it to cope with

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Hi popo,

Thanks for the kind thoughts. The replacement drive didn't come today but as the machine is running happily enough with the degraded array and I'm fully backed up I can live with that until the weekend when I'll have more time. I just hope that the RAID firmware can rebuild the array without booting into an OS as my only access to an OS is by disabling the RAID firmware and setting the disk subsystem to AHCI. If it can't then it will be time to go shopping - unfortunately, if this Xbit comparative review is a good guide, you definitely get what you pay for at the premium end which is very bad news when one sees the price of the Adaptec RAID 5805. But, and it's a very big BUT, that controller will be useful so long as motherboards support PCIe x8 so it might prove a good long term investment. Cover your ears as I tear out what little hair remains on top of my head.

Update: I've been doing a bit more reading and it seems that normally when the Intel Storage Manager is installed a service is also installed which pops up a tray icon which flags a degraded array. It also seems that I should be getting more functionality out of ISM than a simple information pane. So I think the problem is yet again that I'm running the motherboard in a non-standard configuration. The main BIOS only allows me to have a RAID array on the SATA controller while booting off a disk hanging from the SAS controller if I set the disk subsystem to AHCI rather than RAID once I've used RAID to initialise that RAID 10 array. Why it had to be designed like that I have no idea but it may well be that because AHCI is enabled (and so RAID isn't) that the RAID firmware ROM isn't being initialised at system boot up and so the Intel Storage Manager has no way of talking to that same firmware and so expanding it's repertoire. Same deal with the associated service and the missing tray notification. Serves me right for trying to do something ambitious on the cheap.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Hi folks,

The Lian-Li case with it's drive caddies meant that pulling each of the HDDs in turn to check the serial numbers against the known bad drive's serial number was a doddle. Predictably, of course, the unit I needed to replace was the last one pulled.

Reboot into the BIOS to set the disk system as RAID enabled me to access the RAID controller's firmware BIOS and it immediately offered the new drive as the unit with which to rebuild the array. That done it then informed me that the rebuild would occur from within the OS. So a couple of reboots later (needed to reconfigure the main BIOS as documented before) and I'm back in Windows 7 and Intel Storage Manager tells me that everything is hunky-dory.

Unfortunately it's not giving me a progress check on how the rebuild is going (percentage completion) so I'm reliant on the fact that the HDD access light on the front of the case is permanently lit while the rebuild progresses. But I do know from previous experiences of rebuilding SCSI based RAID 5 arrays that this procedure will take many many hours.

The big lesson I've learnt is that due, I assume, to the peculiarities of my set-up I'll not get any warning should a further disk failure occur so I'll have to schedule a weekly look via Intel's Storage Manager. And I'm now resolved to get ...........

OK, while composing this post the computer froze on me for a few seconds and then recovered. Given the way the machine crashed on me a few days ago I immediately went back to ISM and unbelievably the new HDD was also showing "Unknown status" instead of "Normal". In my book that is looking like a controller fault as a cabling fault would not only have showed up immediately but if one existed ISM would be unlikely to be able to report the disk's serial number. Anyway, I opened the machine up again and made sure that the cabling is secure and then put the old drive back in which, from the RAID firmware, had now got flagged as an "Offline member". Removing it from the RAID array via the ICH10R firmware and then adding it back into the RAID 10 array marked it for "Rebuild" which is now proceeding as I'm back into Windows 7.

As I don't have any spare SATA ports I think a further failure indicates dumping the motherboard's controller and maybe going with an Adaptec RAID 5805 or similar. At least I would expect to see the benefit of a rather faster disk system and, as mentioned, that hardware should see service on my next build a few years down the line. The major downside is that I'll also have to do a complete OS and software rebuild of the computer.

Sorry if this saga has proved of little interest but, to paraphrase a famous quiz-master's phrase here in the UK "having started (the story) I had to finish". And maybe it will prove useful to someone else out there on the Net...

Time to get onto the exercise machine and work off some frustration, I think.

Update: This time the rebuild completed normally so perhaps it was an intermittent contact on the SATA cable. Oh well, after some degree of angst I'm now the proud possessor of a spare HDD and I've seriously started to re-evaluate my reliance on motherboard based RAID hardware.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Hi folks,

Just for the record, the RAID 10 array (D: drive) registered another error today, but this time on a different port/disk so the only common factors are now the Intel ICH10R controller on the motherboard, the associated Intel driver and Windows 7 x64. Disk drivers and the associated hardware are fundamental building blocks of any computer and if they can't be trusted they are useless. I've already ordered an Adaptec PCIe SAS/SATA controller to take over from the ICH10R so this latest failure has just moved its installation up the priority list.

Both times when one of these disk errors has occurred the program which I believe was accessing the disk has frozen (Firefox the other day and Outlook just now). Sometimes the freeze releases after many seconds with a "chirrup" from the hard drives. I'm guessing, admittedly, but I think my worst suspicions about the perils of relying on the motherboard's CPU to do the grunt work in managing my RAID 10 array are confirmed.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

At the risk of stating the obvious, given you're going to do a rebuild anyway, probably an idea to scan each disk individually to make absolutely sure they're ok.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.