Disappeared raid

I seem to be having a major problem with my xserve - osx server 10.5.8.
I have my system setup with a terabyte software mirror array and it seems to have just "disappeared" on me. When I cd in terminal to /Volumes/Files, I have two drives showing up (Files9 and Files 1). There obviously should be only one showing up, but for some reason there are two. The /Files volume contains only the directory for my calendar server (/Volumes/Files/CalendarServer/) while the other one contains all my files as is normal (/Volumes/Files 1/). I checked the disk manager for the software raid settings and it checks out okay and verified the disks.
When I browse the drive from Finder, it shows up as "Files", but it points to the correct drive (/dev/disk3 931Gi 784Gi 147Gi 85% /Volumes/Files 1) which is now named wrong.
Why are both of these appearing, why is it not one drive showing up? Any help or suggestions would be greatly appreciated!

Jonathan
At the time you posted this, everything was probably 100% recoverable.
I have no idea what has since transpired, but if you still need assistance, and nothing new has taken place, we can almost certainly help.
Regards
Duncan Clarke
http://www.retrodata.co.uk

Similar Messages

Array recovery?

Xserve RAID with 5x 250Gb Apple disk modules (4 disk RAID 5 array with hot spare) in each side hosted by Windows XP machine with Apple PCI fibre channel card.
Disk 11 (4th in this array) showed red light but the system had auto rebuilt using the hot spare disk 12. I Shutdown the Xserve RAID and removed failed disk. On restart the array was not detected by the Apple fibre channel card BIOS and not visible in Windows. RAID admin sees all the disks in the failed array as spares. Have updated to latest RAID Admin software and controller firmware as this is supposed to be able to re-recognise the array from a set of disks; trouble is the "recognise" button is always greyed out. I have tried swapping the disk sets and controller cards but no change.

Thanks for the reply! I AM using RAID Admin software to manage the array as I stated in my original post. The problem is that the RAID 5 array connected to the bottom RAID controller has disappeared; RAID Admin sees all the disks as spares now. I was hoping that the "Recognise Array" utility in the new RAID Admin V1.5 would be able to recognise the disks.

875P Neo-Fis2R with 2.8c overclock results

With the helpful information that I read from this forum, I've done much tweaking with my system. This is the best I could have so far, and it seems like I can't go further, unless I do something on my hsf, I think. I have 2 more fans, one in front of my casing and another behind. Please see my signature for the general details. Other information is as shown here:
1. Bios settings:
- Ram, 3-4-4-8
- FSB, 232
- PAT, slow
- DOT Disable
- CPU Vcore 1.55V
- DDR Power Voltage 2.70V
- AGP Power Voltage 1.55V
- HT Enabled
2. Changed my Enermax 350W PSU to Pagan 480W (the many molex connectors and SATA connectors, I believe, solved my problem of having sudden reboot of my system, disappearing raid hard disks, etc. The lousy Y-connectors must be the culprit). Not forgetting, maybe the high load coming from my many devices.
3. Sandra 2004 Benchmarks:
CPU Arithemetic Benchmark
- Dhrystone ALU 9726 MIPS
- Whetstone FPU/iSSE2 3929/6971 MFLOPS
Memory Bandwidth Benchmark
- RAM Bandwidth Int Buffered iSSE2 3930 MB/s
- RAM Bandwidth Float Buffered iSSE2 3773 MB/s
Motherboard Information
- Ram 2.0-4-4-8
- PAT Enabled
- Board Temperature 43.0 deg C
- CPU Temperature 45.5 deg C
- Power/Aux Temperature 54 deg C (probably wrong data or un meaningful data)
- Chassis Fan Speed 3448 rpm
- CPU Fan Speed 6618 rpm
Two things strange here:
1. Ram setting in bios different from Sandra Benchmark
2. If Pat is set to other than slow or DOT to enable on any setting, system cannot boot up. Previously, I remembered I could set PAT to at least FAST and DOT to Private or even Captain. Not sure why I can't do it now. Anyway, from what many are suggesting, overclocking the manual way (as what I am doing for my system) is a better way of overclocking.
Help:
1. CPU temperature seems ok now but not sure what it will be if at full load. BTW, anyone can tell me how to do a full load testing.
2. Any other suggestions from anyone, if it is worth it for me to get a better HSF and likely, what kind of improvement can I get.
Thanks.

Hi galf,
Thanks for your suggestions. It turned out that over the weekend, my settings at FSB 232 turns out to be unstable. So got to go back to 230.
FYI, I had set my ram to 2-2-2-6 earlier as well and the benchmarks were about the same as for 3-4-4-8.
3. Sandra 2004 Benchmarks:
CPU Arithemetic Benchmark
- Dhrystone ALU 9617MIPS
- Whetstone FPU/iSSE2 4072/7040 MFLOPS
Memory Bandwidth Benchmark
- RAM Bandwidth Int Buffered iSSE2 4201 MB/s
- RAM Bandwidth Float Buffered iSSE2 4212 MB/s
Anyway, did what you suggested as well.
1. Bios settings:
- Ram, 2-2-2-5, 333 i.e. 5:4 ratio- FSB, 230
- PAT, slow
- DOT Disable
- CPU Vcore 1.55V
- DDR Power Voltage 2.70V
- AGP Power Voltage 1.55V
- HT Enabled
2. Sandra 2004 Benchmarks:
CPU Arithemetic Benchmark
- Dhrystone ALU 9572 MIPS
- Whetstone FPU/iSSE2 4073/7023 MFLOPS
Memory Bandwidth Benchmark
- RAM Bandwidth Int Buffered iSSE2 4645 MB/s
- RAM Bandwidth Float Buffered iSSE2 4669 MB/s
3. Motherboard Information
- Ram 2.5-2-2-5
- PAT Enabled
- Board Temperature 41.0 deg C
- CPU Temperature 46.5 deg C
Tried setting FSB higher but the couldn't boot up. Looks like CPU can only go up to 230 to 232.
So what is better for me, stay with 1:1 or 5:4? Seems like bandwidth is better when on 5:4 but CPU benchmark is better at 1:1. Which is more important, CPU benchmark or memory bandwidth?
Any use to change HSF and if yes, what's a good and reasonable HSF to recommend? Value for money, on the mid range in price, no point to go for low end value for money.
Not sure which ram I have with me, suspect there are not the HyperX models, likely.
Any more suggestions?

Internal RAID Drives Disappeared from Desktop & Finder

A few days ago, one of my two internal mirrored RAID sets no longer showed on the desktop or in the Finder Sidebar. A day later, the other internal RAID set disappeared from desktop and sidebar. The two remaining internal non-RAID drives show up as do two external drives. All internal and external hard drives still show under Disk Utility and Photoshop Bridge CS3.
I checked Finder Preferences and boxes are checked to show internal and external drives on the desktop and sidebar. I looked at the drive detail under Disk Utility and nothing looks wrong with the two RAID sets ON-LINE. I unmounted one RAID set and re-mounted without any change. I tried unchecking the two preference boxes to not show the drives and rebooted, rechecking the boxes and rebooting and nothing changed on the desktop or sidebar.
Does anyone have any suggestions to resolve this?

Well our invisible drives problem is resolved but it's not the solution that I'd hoped for.
We visited the Genius Bar at a local Apple store and they denied any knowledge of problems with invisible drives in Macs and discounted my reports from Mac forums.
I went home and ordered a new 1tb Seagate drive. When it arrived, I installed it in bay 2, copied each invisible drive to the new drive one-at-a-time, broke my two sets of internal mirrored drives, erased all the affected drives and reloaded them. Now all drives are visible on the Desktop and Finder again.
I'm hoping that the disappearing drive problem doesn't surface again but view this as an unresolved Apple issue. Now I can look at other mass storage options for future growth.

Gspeed ES Pro RAID keeps disappearing from Mac, have to re-initialize!

UPDATED VERSION:
MAC PRO KEEPS LOSING CONNECTION TO RAID DRIVE.
After 9 months of organizing a documentary film edit on an external RAID drive (12TB, Raid 0. Yes, I have 2 layers of backup), I purchased the PNY Nvidia Quadro 4000 video card for the primary edit. Soon after, I began having boot up issues. The screen would freeze at various points during the startup process, requiring manual power-downs.
Over the past two months, my computer has lost communication with the RAID 3 times, forcing me to rebuild it from my backup drives (10TB). One tech at Diskwarrior said my computer lost the partition, and that the 1s and 0s had been largely replaced with gibberish.
The Raid becomes unrecognized by my computer in one of two ways:
1- After two days of copying data back over to the raid (rebuilding it), I get an error code -36, saying a file cannot be read or written, then the Raid partition disappears from disk utility and the next time I start up the computer says it doesn’t recognize my drive. The ATTO configuration tool (the card that attaches my raid to the computer) sees the drive, but my computer does not. I have to re-initialize it and start over again.
2- The Raid will go from being recognized to not being recognized when I reboot one day. Again, I have to re-initialize and rebuild.
I removed the PNY Graphics card and put in the stock ATI Radeon HD 5770 it came with. On day 2 of rebuilding the RAID, I got the error -36 and the Raid disappeared from disk utility again. When I opened the raid folder, which seconds before had 10 TB of data on it, it read as zero items. So the PNY Nvidia card was NOT the problem.
Apple has worked on the computer 3 times, running extensive hardware tests, but the same problem persists. At one point they said my blackmagic card was the problem, and more recently they said I probably have bad ram. I tried separate ram (the old 6GB my computer came with), and my computer seemed to begin booting up okay, but one day later I had the -36 error, the RAID couldn’t be recognized again and I lost the data during another rebuild. Apple even replaced the Ram Tray, but the same problem persists.
G technology says my RAID’s hardware is fine, and ATTO, who makes the Raid card, says it looks fine to them too. This does not mean they are 100% not the problem though.
Yesterday as I was attempting to rebuild the RAID again, I saw the -36 error come up, and files stopped copying. I immediately made a screenshot and noted the time.
Today I spoke with a really cool tech at Apple named Ron who helped me investigate the Kernel log, and this is what it said (the RAID is called FadingWest):
Dec 6 17:05:36 Jesses-Mac-Pro kernel[0]: hfs_swap_BTNode: invalid forward link (0xbb739311 >= 0x0000df00)
Dec 6 17:05:36 Jesses-Mac-Pro kernel[0]: hfs: node=2717 fileID=4 volume=FadingWest device=/dev/disk3s2
Dec 6 17:05:36 Jesses-Mac-Pro kernel[0]: hfs: Runtime corruption detected on FadingWest, fsck will be forced on next mount.
Ron believes my problem is with the RAID card. He thinks it’s dropping the connection. So I’ve got an RMA going with ATTO for a new one.
Here are the final things I can troubleshoot at this point:
New RAID card (ATTO)
New RAID drive (G-Technology)
Motherboard / PCI lanes (Apple).
NEW COMPUTER (Apple)
Any ideas are appreciated. Here are the items in question:
Mac Pro mid 2010 2x2.66 Ghz 6 core Intel Xeon OS X Lion 10.7.5
G speed ES Pro (G Technology) – 12TB @ RAID 0
ATTO R680 RAID card

I found this on another help site, I think it's the trick:
there are 2 halves to the Dock. With a bar in the middle.
on the left side, there are 2 sections. the furthest left are the icons that will stay in the dock, then on the right side of the left half of the dock, are the temporary icons, the ones that just stay there till the program is done.
You can freely move a temporary icon to the left and it will be come a permanent icon.
I am reminded of the wizard of OZ here... you could always get to firefox... you just had to click your mouse (and drag an icon)..

RAID disappearing after binding to Directory Service

We have an Xserve (late 2006) with Leopard Server. I built and RAID 1. After binding this server to a directory service, master running on a 10.7.5 server (Mac mini) the RAID is disappearing. If I click on the volume I get an error -1407.
After unbinding the Xserve from DS and mounting the RAID with disk utility everything is fine again.
Background:
Want to use the RAID on the Xserve for homes and group directories of the users.
Please help, I don't know what to do further....
Thanks in advance
Jens

Had to unbind from AD. and just log on locally.

G-Raid drive disappeared while using FCP. Will no longer mount.

Ah! Help! I'm on deadline to edit a film trailer for tomorrow and my hard drive and all my footage just disappeared. I was using Final Cut Pro (6.0.6?) with my G-RAID 2TB drive as my scratch disc, and while I was using it, the drive unmounted. I tried powering it off. Restarting the computer. Swapping out new fw800 cables. Switching to a differnt fw800 port on my drive. Restarting again. The drive powers up but does not show. I then tried putting the fw800 cables into my other smaller drive. This also did not show up. I've used two sets of cables. I don't have any other cables to try. But it makes me feel like something is wrong with the Mac now. I am very tech inept and stressed about this deadline so any help would be great. I recently (a few weeks ago) upgraded to 10.5.8 and have had no problems. (this was after an intital upgrade several months ago that worked fine until all of a sudden it got stuck on the login page. I reset the pram and pressed C and and a number of things on start up and eventually rebooted from the leopard discs and reinstalled with no problems. just mentioning it in case). From time to time when I have started up FCP in the past, it has not been able to see my g-raid drive, but that just seemed to be a factor of me starting it up too late or out of order.
Imac 7,1
Intel Core 2 Duo
10.5.7
Any more info I can provide? Any help would be great! I am desperate to have my footage back and computer working so that i can make this grant deadline tomorrow!
Thank you!

Try doing a SMC reset, that might do it. Make sure ALL peripherals are unplugged when you do it and plug them one at time, test, next one, test etc...
_SMC RESET_
• Shut down the computer.
• Unplug the computer's power cord and all peripherals.
• Press and hold the power button for 5 seconds.
• Release the power button.
• Attach the computers power cable.
• Press the power button to turn on the computer.
Roger

Left side of XServe Raid disappeared. Please help

Hello,
During the cleanup of a bunch of data on right side of my XServe Raid (two raid 5's with spare, one on the left side, one on the right side), the server became very unresponsive (Windows 2k server), to the point where it wouldn't even shut down. I reset the box but when it came back up the left side of the XServe was gone (showed up as a missing drive in disk manager).
RAID Admin shows that there is no raid on the left side and shows all drives as spares. I've just completed copying all the data off of the still functional right hand side and can now attempt to recover the left side (if that's even possible?). I saw a suggestion that full power offs and restarts (multiple times) has helped another poster. So far I have powered down the XServe Raid 2 or 3 times without the left side raid showing up. I'll be trying a few more times shortly.
I believe there is an option to recover the raid in RAID Admin? My XServe has an older version of firmware that does not support the RAID Admin Utilities. I have been nervous of upgrading the firmware and possibly irreversibly losing the raid if it isn't already irreversibly gone.
Can anyone offer any suggestions?

No luck on the reset, both via Raid Admin and the reset button on the XServe Raid.
A restore of most of the data has completed onto another storage device.
I can see a power spike or something wiping out the config of both arrays, but to have only one side just disappear and with no way to re-read the raid configuration from the drives, I can't trust the XServe RAID anymore and we won't be going back to using it.
I certainly appreciate you taking the time to offer a suggestion Camelot!

Volumes and raid sets disappeared

I booted my machine and it gave me a question mark folder I booted it off an external drive but the internal bay hdds arent mounting. When I was looking through raid utility it says they are all good and viable disks in roaming but no raid sets or volumes are there anymore. This all happened to my xserve (2008 2.8ghz 4gb ram) Three apple brand hdd at 1TB(in a raid 5 configuration) the os I had installed on the external hdd was 10.5.3..... any ideas or thoughts to getting the raid sets or volumes back.... also Disk utility still doesnt recogonize the three internal bays is that cause for concern or do you have to have a raid set and volume before disk utility to recogonize it. Thank you for your time and insight to my problem!

Use this guide:
RAID Utility - User Guide
To delete a RAID set or a volume:
1 Open RAID Utility and choose the volume or RAID set in the left column.
2 Click Delete Volume or Delete RAID Set.

Memory Disappearing after adding Raid Controller

I have a Lenovo D20 with 2x x5650 CPUs. I have 9 sticks of 2gb 1333 ECC UDIMM.
The problem I am running into is that after I installed a hardware Raid card, the system only picks up 12gb of memory instead of the installed 18gb, and if i take the card out it'll go back to 18gb.
I've updated to the latest BIOS and still doesn't make a difference. Does anyone have any idea why?

Which specific slots do you think are not workign correctly (ID the slots per the silkscreen on the motherboard)? Watching the system count memory won't tell you this, only swapping DIMMs in and out will.
Can you provide the vendor part number for the DIMMs you are using?
What happens when you put memory only into the DIMM slot(s) you suspect aren't working and try to boot the system? Does it boot or does it give you a beep code sequence?
I can't think of any reason why installing a PCIe device would affect the amount of memory being detected within BIOS. Cards can, however, greatly affect the amount of memory that shows available within the OS.
Often, when there's a mismatch between the amount of physically installed memory and the amount of memory being detected in BIOS (in the BIOS setup menus), it's because of DIMM(s) that are not seated correctly. Obviously the latches should be closed, but sometimes even when closed I've seen a DIMM not report correctly. When installing/reseating, try pushing on the center of the DIMM evenly instead of out near the corners.

In Design / Mac OS 10.8 / using External RAID / shortcuts disappear

Friends,
When I use In Design, the key to my success is using a whole bunch of shortcuts in my FINDER WINDOW:
But then, when I unplug my RAID external drive from my laptop, I lose all my shortcuts!!!!
Like this:
What can I do to fix this??

I have posted on the Mac forum and absolutely nobody has responded:
https://discussions.apple.com/thread/4898631

Raid controller disappears from BIOS

Hi,
I got a strange problem: everytime I pull out the power cord from the computer-the Raid controller disappers from Bios and of course the comp. cant find the HDD`s.
After some time I found out how to get it work again,
I have to pull out the IDE cable and power cable from the HDD and put it back again. Then the Fast Track Controller comes up again in BIOS.
Any one has an idea what that could be?
I use The Promise Fast Trak 100 Raid driver.
Thanks for reply.

Quote
Originally posted by Bas
Hi,
You should use the MSI raid drivers....
Also, check the jumper on your harddisk and the way the cables are connected, the blue connector should be on the raid controller...
What harddisks do you use?
Hi Bas,
Thanks for reply, I use 2x IBM 60GXP, 60GB HDD.
My drivers are from MSI.
Every thing worked fine almost 1 year, I did not change anything.
First I thougt maybe it could be a battery the voltage is 3.41V as it was when I bought the mobo.

400GB Raid partition disappear -after file transfer from USB2 HD

Hallo, I have now been using my system prodly for two months, or so. But under a filetransfer from my removable USB 2.0 HD, a disaster happened An error message that said; Faulty writing. And my 400GB was gone from Windows
I have checked the raid drivers, and reinstalled them, and I can see at start up that the raid array is functional. With Seagate tool I can see that there is a partition of 400GB. What has appened
Anyone I have over 200GB of Movies on that partition, any solution on this
How to make Windows see the raid array partiton agin
I use still winXP SP2

Sorry for the late feedback, think I had to repress it for a while, the los of 200Gb
I used stripe Raid 0 for better speed...
Now I have given up Raid, so Iam using the IDE 3 & 4 as regular IDE connection for HD's.
So now I have started from scratch again .

[SOLVED] mdadm / RAID trouble

Hi all,
In relation to this post: Migrating data to a "new" setup, questions
I think I've a problem with my new RAID5 array everything seemed to work fine, I was transferring stuff to it, which went ok. Then I reduced the old fs on /home (which held everything) to the minimum size ~2.3TB, and pvmove:ed a partition off it so I could add it to the RAID array. After unmounting the raid to do an fsck, something happened (though it didn't say!), it's like the RAID array just disappeared.
Here's the dmesg: http://dpaste.org/1oUi/ snippet below:
md0: detected capacity change from 4000795590656 to 0
md: md0 stopped.
md: unbind<sdb>
md: export_rdev(sdb)
md: unbind<sda>
md: export_rdev(sda)
md: bind<sda1>
md: bind<sdb1>
md: bind<sdc1>
md/raid:md0: device sdb1 operational as raid disk 1
md/raid:md0: device sda1 operational as raid disk 0
md/raid:md0: allocated 3175kB
md/raid:md0: raid level 5 active with 2 out of 3 devices, algorithm 2
RAID conf printout:
--- level:5 rd:3 wd:2
disk 0, o:1, dev:sda1
disk 1, o:1, dev:sdb1
md0: detected capacity change from 0 to 4000793231360
RAID conf printout:
--- level:5 rd:3 wd:2
disk 0, o:1, dev:sda1
disk 1, o:1, dev:sdb1
disk 2, o:1, dev:sdc1
md: recovery of RAID array md0
md: minimum _guaranteed_ speed: 1000 KB/sec/disk.
md: using maximum available idle IO bandwidth (but not more than 200000 KB/sec) for recovery.
md: using 128k window, over a total of 1953512320 blocks.
md0: detected capacity change from 0 to 4000793231360
md0: unknown partition table
EXT4-fs (dm-2): mounted filesystem with ordered data mode. Opts: (null)
EXT4-fs (dm-3): mounted filesystem with ordered data mode. Opts: (null)
EXT4-fs (dm-2): re-mounted. Opts: (null)
ata2.00: exception Emask 0x0 SAct 0x7ff SErr 0x0 action 0x6 frozen
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/00:00:00:cf:70/01:00:e7:00:00/40 tag 0 ncq 131072 in
res 40/00:04:00:37:4b/00:00:de:00:00/40 Emask 0x4 (timeout)
etc etc...
Here is messages.log: http://dpaste.org/OO7I/
I also have the SMART (smartctl --all) info for all (sda ... sdg) drives:http://dpaste.org/40Wd/
I'm rebooting the server now to see if that helps...
edit: Ugh, can't ssh in! BALLS. I'll have to check the status when I get home..
In the mean time, do you guys think the array is borked?
Last edited by Fackamato (2010-11-01 12:10:55)

This is weird. The array was recovering during the night, and today I get this:
ata2: EH complete
ata2.00: NCQ disabled due to excessive errors
ata2.00: exception Emask 0x0 SAct 0x7fe SErr 0x0 action 0x6 frozen
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/00:08:00:d0:70/01:00:e7:00:00/40 tag 1 ncq 131072 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/38:10:c8:cf:70/00:00:e7:00:00/40 tag 2 ncq 28672 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/c8:18:00:cf:70/00:00:e7:00:00/40 tag 3 ncq 102400 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/00:20:00:ce:70/01:00:e7:00:00/40 tag 4 ncq 131072 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/e8:28:18:cd:70/00:00:e7:00:00/40 tag 5 ncq 118784 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/10:30:08:cd:70/00:00:e7:00:00/40 tag 6 ncq 8192 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/00:38:00:d2:70/01:00:e7:00:00/40 tag 7 ncq 131072 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/08:40:00:cd:70/00:00:e7:00:00/40 tag 8 ncq 4096 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/00:48:00:cc:70/01:00:e7:00:00/40 tag 9 ncq 131072 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2.00: failed command: READ FPDMA QUEUED
ata2.00: cmd 60/00:50:00:d1:70/01:00:e7:00:00/40 tag 10 ncq 131072 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2: hard resetting link
ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata2.00: configured for UDMA/133
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2.00: device reported invalid CHS sector 0
ata2: EH complete
ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
ata2.00: failed command: READ DMA EXT
ata2.00: cmd 25/00:00:00:cc:70/00:01:e7:00:00/e0 tag 0 dma 131072 in
res 40/00:0c:a8:6c:e9/00:00:e1:00:00/40 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2: hard resetting link
ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata2.00: configured for UDMA/133
ata2: EH complete
ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
ata2.00: failed command: READ DMA EXT
ata2.00: cmd 25/00:00:00:cc:70/00:01:e7:00:00/e0 tag 0 dma 131072 in
res 40/00:0c:a8:6c:e9/00:00:e1:00:00/40 Emask 0x4 (timeout)
ata2.00: status: { DRDY }
ata2: hard resetting link
ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata2.00: configured for UDMA/133
sd 1:0:0:0: [sdb] Result: hostbyte=0x00 driverbyte=0x08
sd 1:0:0:0: [sdb] Sense Key : 0xb [current] [descriptor]
Descriptor sense data with sense descriptors (in hex):
72 0b 00 00 00 00 00 0c 00 0a 80 00 00 00 00 00
00 00 00 a7
sd 1:0:0:0: [sdb] ASC=0x0 ASCQ=0x0
sd 1:0:0:0: [sdb] CDB: cdb[0]=0x28: 28 00 e7 70 cc 00 00 01 00 00
end_request: I/O error, dev sdb, sector 3882929152
md/raid:md0: read error not correctable (sector 3882927104 on sdb1).
md/raid:md0: Disk failure on sdb1, disabling device.
<1>md/raid:md0: Operation continuing on 1 devices.
md/raid:md0: read error not correctable (sector 3882927112 on sdb1).
md/raid:md0: read error not correctable (sector 3882927120 on sdb1).
md/raid:md0: read error not correctable (sector 3882927128 on sdb1).
md/raid:md0: read error not correctable (sector 3882927136 on sdb1).
md/raid:md0: read error not correctable (sector 3882927144 on sdb1).
md/raid:md0: read error not correctable (sector 3882927152 on sdb1).
md/raid:md0: read error not correctable (sector 3882927160 on sdb1).
md/raid:md0: read error not correctable (sector 3882927168 on sdb1).
md/raid:md0: read error not correctable (sector 3882927176 on sdb1).
ata2: EH complete
md: md0: recovery done.
etc... Hm. Is sdb borked? How can I check it?
edit: Trying to recover.
[root@ion ~]# cat /proc/mdstat && mdadm --detail --verbose /dev/md0
Personalities : [raid6] [raid5] [raid4] [raid0]
md0 : active raid5 sda1[0] sdc1[3] sdb1[1]
3907024640 blocks super 1.2 level 5, 128k chunk, algorithm 2 [3/2] [UU_]
[=>...................] recovery = 6.9% (134951552/1953512320) finish=429.1min speed=70626K/sec
unused devices: <none>
/dev/md0:
Version : 1.2
Creation Time : Tue Oct 19 08:58:41 2010
Raid Level : raid5
Array Size : 3907024640 (3726.03 GiB 4000.79 GB)
Used Dev Size : 1953512320 (1863.01 GiB 2000.40 GB)
Raid Devices : 3
Total Devices : 3
Persistence : Superblock is persistent
Update Time : Thu Oct 21 09:05:13 2010
State : clean, degraded, recovering
Active Devices : 2
Working Devices : 3
Failed Devices : 0
Spare Devices : 1
Layout : left-symmetric
Chunk Size : 128K
Rebuild Status : 6% complete
Name : ion:0 (local to host ion)
UUID : e6595c64:b3ae90b3:f01133ac:3f402d20
Events : 23550
Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 8 17 1 active sync /dev/sdb1
3 8 33 2 spare rebuilding /dev/sdc1
I tried to add the forth drive (sdd1) but couldn't:
mdadm –add /dev/md0 /dev/hdd1
mdadm: add new device failed for /dev/hdd1 as 4: Invalid argument
edit: That's because the raid was active, duh. Stopped md0, could add the drive. I couldn't _grow_ it though, as it's recovering..
[root@ion ~]# mdadm -D /dev/md0
/dev/md0:
Version : 1.2
Creation Time : Tue Oct 19 08:58:41 2010
Raid Level : raid5
Array Size : 3907024640 (3726.03 GiB 4000.79 GB)
Used Dev Size : 1953512320 (1863.01 GiB 2000.40 GB)
Raid Devices : 3
Total Devices : 4
Persistence : Superblock is persistent
Update Time : Thu Oct 21 12:41:26 2010
State : clean, degraded, recovering
Active Devices : 2
Working Devices : 4
Failed Devices : 0
Spare Devices : 2
Layout : left-symmetric
Chunk Size : 128K
Rebuild Status : 52% complete
Name : ion:0 (local to host ion)
UUID : e6595c64:b3ae90b3:f01133ac:3f402d20
Events : 23568
Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 8 17 1 active sync /dev/sdb1
3 8 33 2 spare rebuilding /dev/sdc1
4 8 49 - spare /dev/sdd1
[root@ion ~]# cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4] [raid0]
md0 : active raid5 sdd1[4](S) sda1[0] sdc1[3] sdb1[1]
3907024640 blocks super 1.2 level 5, 128k chunk, algorithm 2 [3/2] [UU_]
[==========>..........] recovery = 52.0% (1016981760/1953512320) finish=232.6min speed=67084K/sec
unused devices: <none>
Last edited by Fackamato (2010-10-21 10:42:18)

Mac Pro RAID 5 disk goes offline frequently in a random way. What can be causing this odd behavior?

Hi dudes,
I installed recently two Mac Pro RAID cards inside their corresponding Mac Pro systems. Four 2 TB Hitachi SATA disks are contained inside each system, and configured as RAID 5. Yes, the operating system is installed on this RAID 5 volume in order to get the highest performance of the array, taking advantage of the RAID 5 protection (Autodesk Smoke runs in each system; this application works only with uncompressed video in real time). I already had tested both systems keeping separated the boot disk just for the operating system, and making the RAID 5 volume just with three 2 TB disks, but the performance was slow on both systems.
The performance is now very good, but unfortunately, in both systems happens that sometimes one of the disks goes offline with no apparent reason. The RAID Utility immediately reports the failure and it is mandatory to declare as spare the disk that went offline. Then the rebuild process begins, but in the meantime, the performance goes down in a noticeable way. Sometimes it is even worse, because the disk disappears completely. Then, it is mandatory to turn off the system, and boot it up again, in order to see in the RAID Utility the missing disk, which needs to be declared as spare in order to be reintegrated into the RAID 5 volume in a slow rebuilding process.
Some important remarks: this very new Mac Pro systems do not have the iPass cable (at least apparently; I already disassembled completely one of these systems). This cable is mentioned in one Mac RAID card manual that I found over the internet. The diagrams do not match exactly with the Mac Pro nor the Mac Pro RAID board. I did not find a proper iPass connector in any of the Mac Pro RAID cards (?!). So, my guess is that currently the communication between the system and its Mac Pro RAID card is just by means of the internal bus. I think that if the iPass connection were mandatory, the RAID Utility would report it with a noticeable error message. Please advise.
I think that these Mac Pro RAID cards need a firmware upgrade, in order to be able to work fine with big SATA drives. Again, please advise.
Thanks in advance from Mexico.
Sincerely,
Martin Ponce de Leon

Hi Grant,
Thank you for your prompt response. The manual is a PDF document issued by Apple and it seems OK, but no updated to the latest Mac Pro system and Mac Pro RAID. I do not find the link where I found this manual. The systems belong to one of our customers. As far as I remember, in the printed manual included with the boards, it is not mentioned anything about the iPass cable, just about the battery cable. Do you know where can I get a PDF manual of the latest Mac Pro RAID card and the latest Mac Pro system? This client is far from our offices. So, I would prefer a PDF copy of this manual.
The drives that have gone offline, once they are back, are reported in a good status in the RAID Utility. Besides that, I have tested them in another system in our facilites, and all of them work fine. So, my guess is that the high capacity drives are not yet supported by the Mac Pro RAID card, or it requires a firmware update.
They do need to have in the same volume OS and video storage because three disks do no provide a good performance (they need also RAID 5), but four disks work fine... excepting when one disk is missing. Please advise. Thank you.
Best Regards,
Martin

Disappeared raid

Similar Messages

Maybe you are looking for