August 31, 201015 yr Author If running with the disk assigned to the array, you must use the /dev/md1 device If running reiserfsck on the raw disk partition, you must use partition1 ( /dev/sdc1 ). reiserfsck --check /dev/sdc1 (note the trailing 1 indicating the first partition) You can fix the HPA issue after you get the array back to a non-degraded state, but before you do this, you'll need to upgrade the BIOS on your MB. Thanks for the help =) Running "reiserfsck --check /dev/sdc1" took a while and returned saying everything was fine. It's still coming up as unformatted though. I also updated the bios on my motherboard to the latest and turned off the option to backup the bios to the HDD (which I assume is the feature causing the problem). How should I proceed?
August 31, 201015 yr If running with the disk assigned to the array, you must use the /dev/md1 device If running reiserfsck on the raw disk partition, you must use partition1 ( /dev/sdc1 ). reiserfsck --check /dev/sdc1 (note the trailing 1 indicating the first partition) You can fix the HPA issue after you get the array back to a non-degraded state, but before you do this, you'll need to upgrade the BIOS on your MB. Thanks for the help =) Running "reiserfsck --check /dev/sdc1" took a while and returned saying everything was fine. It's still coming up as unformatted though. I also updated the bios on my motherboard to the latest and turned off the option to backup the bios to the HDD (which I assume is the feature causing the problem). How should I proceed? OK, that is a good first step, although as many have mentioned if the "default" is to back-up to HDD, then you'll get a surprise one morning when the CMOS battery dies and the BIOS reverts back to its defaults. Simply checking the file-system on the disk will not mount it. the easiest way to get it mounted is to simply press the "Stop" button on the management web-interface. then press the "Start" button. If that does not do it, then stop the array and reboot.
August 31, 201015 yr Author If running with the disk assigned to the array, you must use the /dev/md1 device If running reiserfsck on the raw disk partition, you must use partition1 ( /dev/sdc1 ). reiserfsck --check /dev/sdc1 (note the trailing 1 indicating the first partition) You can fix the HPA issue after you get the array back to a non-degraded state, but before you do this, you'll need to upgrade the BIOS on your MB. Thanks for the help =) Running "reiserfsck --check /dev/sdc1" took a while and returned saying everything was fine. It's still coming up as unformatted though. I also updated the bios on my motherboard to the latest and turned off the option to backup the bios to the HDD (which I assume is the feature causing the problem). How should I proceed? OK, that is a good first step, although as many have mentioned if the "default" is to back-up to HDD, then you'll get a surprise one morning when the CMOS battery dies and the BIOS reverts back to its defaults. Simply checking the file-system on the disk will not mount it. the easiest way to get it mounted is to simply press the "Stop" button on the management web-interface. then press the "Start" button. If that does not do it, then stop the array and reboot. Ok. I hit stop and then start but after saying "mounting" on the dodgy disk1 for a few seconds went back to unformatted and red. When I rebooted it still said the same thing. Seems strange that the check tool can see all my files (i see them pop up when its checking) but unraid sees it as unformatted... is there perhaps a way (if there isnt a way to get unraid to see it...) i can mount it myself outside unraid and copy to another drive in the array?
August 31, 201015 yr If running with the disk assigned to the array, you must use the /dev/md1 device If running reiserfsck on the raw disk partition, you must use partition1 ( /dev/sdc1 ). reiserfsck --check /dev/sdc1 (note the trailing 1 indicating the first partition) You can fix the HPA issue after you get the array back to a non-degraded state, but before you do this, you'll need to upgrade the BIOS on your MB. Thanks for the help =) Running "reiserfsck --check /dev/sdc1" took a while and returned saying everything was fine. It's still coming up as unformatted though. I also updated the bios on my motherboard to the latest and turned off the option to backup the bios to the HDD (which I assume is the feature causing the problem). How should I proceed? OK, that is a good first step, although as many have mentioned if the "default" is to back-up to HDD, then you'll get a surprise one morning when the CMOS battery dies and the BIOS reverts back to its defaults. Simply checking the file-system on the disk will not mount it. the easiest way to get it mounted is to simply press the "Stop" button on the management web-interface. then press the "Start" button. If that does not do it, then stop the array and reboot. Ok. I hit stop and then start but after saying "mounting" on the dodgy disk1 for a few seconds went back to unformatted and red. When I rebooted it still said the same thing. Seems strange that the check tool can see all my files (i see them pop up when its checking) but unraid sees it as unformatted... is there perhaps a way (if there isnt a way to get unraid to see it...) i can mount it myself outside unraid and copy to another drive in the array? Is there perhaps a way you'll post a syslog now, so we'll see the real reason for the "red" indication. For that "red" indicator to occur, "writes" to the drive must have failed, but the only way to know what is happening is to see the syslog.
August 31, 201015 yr Author If running with the disk assigned to the array, you must use the /dev/md1 device If running reiserfsck on the raw disk partition, you must use partition1 ( /dev/sdc1 ). reiserfsck --check /dev/sdc1 (note the trailing 1 indicating the first partition) You can fix the HPA issue after you get the array back to a non-degraded state, but before you do this, you'll need to upgrade the BIOS on your MB. Thanks for the help =) Running "reiserfsck --check /dev/sdc1" took a while and returned saying everything was fine. It's still coming up as unformatted though. I also updated the bios on my motherboard to the latest and turned off the option to backup the bios to the HDD (which I assume is the feature causing the problem). How should I proceed? OK, that is a good first step, although as many have mentioned if the "default" is to back-up to HDD, then you'll get a surprise one morning when the CMOS battery dies and the BIOS reverts back to its defaults. Simply checking the file-system on the disk will not mount it. the easiest way to get it mounted is to simply press the "Stop" button on the management web-interface. then press the "Start" button. If that does not do it, then stop the array and reboot. Ok. I hit stop and then start but after saying "mounting" on the dodgy disk1 for a few seconds went back to unformatted and red. When I rebooted it still said the same thing. Seems strange that the check tool can see all my files (i see them pop up when its checking) but unraid sees it as unformatted... is there perhaps a way (if there isnt a way to get unraid to see it...) i can mount it myself outside unraid and copy to another drive in the array? Is there perhaps a way you'll post a syslog now, so we'll see the real reason for the "red" indication. For that "red" indicator to occur, "writes" to the drive must have failed, but the only way to know what is happening is to see the syslog. Point taken, I'll stick to reporting what I see =) Syslog is attached. syslog-2010-08-31.txt
August 31, 201015 yr I see no failing "writes" in this syslog, therefore I'm going to assume they were from before you rebooted and had disk1 un-assigned. The disk is being emulated, and since you did NOT run the reiserfsck on the /dev/md1 device, but instead on the actual physical device, it does not see the repairs you made. You can probably now initialize the disk configuration, and then tell unRAID that the parity is OK (even though we know it is NOT) and then let it perform a full parity check where it will find differences and get parity back in sync with the disks themselves. So... To do this you need to "Stop" the array Then, log into the server via telnet or on the system console and type these two commands: /root/initconfig It will prompt you to be certain the array is stopped, respond with "Yes" (Upper case "Y", lower case "es") Then, type this command: /root/mdcmd set invalidslot 99 Then, back on the web-interface, press the "refresh" button on the browser. All the indicators should show as blue. Lastly, press "Start" and let the parity check run to completion. It will find parity errors... do not get alarmed when you see them as you caused them when running reiserfsck on the /dev/sdc1 device instead of the /dev/md1 device.
September 1, 201015 yr Author I see no failing "writes" in this syslog, therefore I'm going to assume they were from before you rebooted and had disk1 un-assigned. The disk is being emulated, and since you did NOT run the reiserfsck on the /dev/md1 device, but instead on the actual physical device, it does not see the repairs you made. You can probably now initialize the disk configuration, and then tell unRAID that the parity is OK (even though we know it is NOT) and then let it perform a full parity check where it will find differences and get parity back in sync with the disks themselves. So... To do this you need to "Stop" the array Then, log into the server via telnet or on the system console and type these two commands: /root/initconfig It will prompt you to be certain the array is stopped, respond with "Yes" (Upper case "Y", lower case "es") Then, type this command: /root/mdcmd set invalidslot 99 Then, back on the web-interface, press the "refresh" button on the browser. All the indicators should show as blue. Lastly, press "Start" and let the parity check run to completion. It will find parity errors... do not get alarmed when you see them as you caused them when running reiserfsck on the /dev/sdc1 device instead of the /dev/md1 device. That syslog is current, from the machine as it is right now after rebooting. I've attached a screenshot too. Are you sure the drive is being emulated? The parity shouldn't be valid as it was before i took two hard drives out. I tried running "reiserfsck --check /dev/md1" but it said The problem has occurred looks like a hardware problem... and bread: Cannot read the block (2): (Input/output error). Aborted And in the syslog it added Sep 1 00:04:44 Destructor kernel: Buffer I/O error on device md1, logical block 2 Sep 1 00:04:44 Destructor kernel: Buffer I/O error on device md1, logical block 2 Sep 1 00:05:47 Destructor ata_id[2792]: HDIO_GET_IDENTITY failed for '/dev/block/8:32' Sep 1 00:05:47 Destructor ata_id[2816]: HDIO_GET_IDENTITY failed for '/dev/block/8:48' Sep 1 00:05:47 Destructor ata_id[2831]: HDIO_GET_IDENTITY failed for '/dev/block/8:64' Sep 1 00:05:47 Destructor ata_id[2844]: HDIO_GET_IDENTITY failed for '/dev/block/8:80'
September 3, 201015 yr Author Hooray its all fixed!!! I took the drive acting up out then did "initconfig" and started the array. I cancelled the parity check, put the drive back in, assigned it to slot one and started it up again! Hooray! "Unformatted" threw me a little bit. It seems the drive was formatted just fine? Perhaps there's a better word for it (or far more likely I just dont really understand what was going on). Parity is being rebuilt now. Thankyou very much for the help all =) I now have a motherboard that wont destroy my drives (unless cmos loses power ) and a slightly better understanding of the tech I rely on to keep all my media in order. Cheers! Elliot.
September 3, 201015 yr Author I was wrong... so very very wrong. Half way through the parity check... loads of errors on disk 1. Stopped parity check. Now "unformatted". Back to where I was.
September 5, 201015 yr Author A happy ending. I added the drive again but this time didnt start a parity check. I unassigned the parity (to speed up transfer speed) and copied everything from disk1 to disk5. Got everything off without any errors. Took out disk1, put parity back, initconfig'd and started the parity rebuild. Running fine now... not sure whats wrong with the drive so I'll run diagnostics before sending it back to the manufacturer. =)
Archived
This topic is now archived and is closed to further replies.