Okay,

 

Thought I'd start a new topic so my problem didn't get lost in the other.    I have a 1TB EARS drive that was fine in unRAID aside from really slow writes.  I read the thread about applying a jumper to pins 7/8 so I did.    unRAID wouldn't format it and gave 2 errors.  I then started a preclear and cancelled it shortly after it starting.  The drive then was allowed to be added to the array and a re-build started.  At the end of it all it finished but wrote nothing to the disk.  My shares aren't there and no data.  I am puzzled to say the least.  I have my data backed up on a few drives I have laying around but what gives.

 

Is it because I start/stopped the preclear script?  After formatting the disk as NTFS and as ext2 and re-adding them once more the option to format doesn't appear it just says that it's rebuilding the array and of course doesn't actually do it although it runs it's course.

 

Any ideas?  I am performing a preclear with the -n designation right now to see if cancelling the initial preclear somehow messed it up and unRAID is somehow thinking it's all set to go.  Strangely the -t returns it as not precleared.

 

Any faster way to get this show on the road?

 

Kryspy

Adding the jumper to the drive changes the sector alignment and effectively destroys any data that was on it. http://lime-technology.com/wiki/index.php?title=Advanced_Format_Drives

 

As long as you have valid parity drive, you should be able to rebuild it.

 

Since you're going to be trusting this server with your data, its probably not in your best interest to look for quick and dirty ways to get it up and running.  A full preclear (with the jumper installed) is very useful for testing the disk as well as discovering any issues.  The  pre and post run SMART reports will also aid in determining whether the disk is in good shape and ready to be used.

 

Personally I would run a full preclear on it.  It could take upwards of 20 hours, but in the grand scheme of things I'd rather have some assurance that everything is in good shape.

 

Then, assuming that you have a valid parity drive, replace the missing drive in the array with this one and let it rebuild the data.  Again, that will take some time, but you can use it while this is occurring.

Hi,

 

I had run a full preclear on this drive previously so that is why I am running the -n option now.

 

Kryspy

With the drive out of the array, are you able to access the data on the emulated disk?  For example, if your problematic drive was disk1 in the array, when you removed it from the array you should have still been able to access the data on disk1 (as it would be emulated by the parity + all the other data disks).

No I was not.    I assume that means that parity was invalid then?   If I have to redo this I guess I will switch around the EARS drive and make it parity and use the other as disk1 while I am at it.    I should add that I am only running with a disk1 and parity.  Do I have to have run with 3 disks for this to work?

 

Kryspy

Something is odd here, post a syslog.  Did you see any colors besides green next to any of the other drives (besides the problematic one)?  Have you been running regular parity checks?

 

Also, since your shares are missing, it is possible that your flash drive died or became corrupted.  Have you had any boot issues?  Are you able to access the flash share?

Syslog attached.   Also, flash disk is accessible along with disk1.   Only problem is disk1 is empty.  No boot issues.  Only issue was me deciding to fool around with the pin out on the drive

 

Kryspy

syslog-2010-10-13.txt

Don't leave me hangin' 

 

Kryspy

At the start of the syslog I can see /dev/sda being identified

Oct 13 08:56:21 Server kernel: ata1.00: ATA-8: WDC WD10EADS-00M2B0, 01.00A01, max UDMA/133

Oct 13 08:56:21 Server kernel: ata1.00: 1953525168 sectors, multi 16: LBA48 NCQ (depth 0/32)

Oct 13 08:56:21 Server kernel: ata1.00: configured for UDMA/133

Oct 13 08:56:21 Server kernel: scsi 1:0:0:0: Direct-Access    ATA      WDC WD10EADS-00M 01.0 PQ: 0 ANSI: 5

Oct 13 08:56:21 Server kernel: sd 1:0:0:0: [sda] 1953525168 512-byte logical blocks: (1.00 TB/931 GiB)

Oct 13 08:56:21 Server kernel: sd 1:0:0:0: [sda] Write Protect is off

Oct 13 08:56:21 Server kernel: sd 1:0:0:0: [sda] Mode Sense: 00 3a 00 00

Oct 13 08:56:21 Server kernel: sd 1:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA

Oct 13 08:56:21 Server kernel:  sda: sda1

 

And then /dev/sdb

Oct 13 08:56:21 Server kernel: ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Oct 13 08:56:21 Server kernel: ata2.00: ATA-8: WDC WD10EARS-00Y5B1, 80.00A80, max UDMA/133

Oct 13 08:56:21 Server kernel: ata2.00: 1953525168 sectors, multi 16: LBA48 NCQ (depth 0/32)

Oct 13 08:56:21 Server kernel: ata2.00: configured for UDMA/133

Oct 13 08:56:21 Server kernel: scsi 2:0:0:0: Direct-Access    ATA      WDC WD10EARS-00Y 80.0 PQ: 0 ANSI: 5

Oct 13 08:56:21 Server kernel: sd 2:0:0:0: [sdb] 1953525168 512-byte logical blocks: (1.00 TB/931 GiB)

Oct 13 08:56:21 Server kernel: sd 2:0:0:0: [sdb] Write Protect is off

Oct 13 08:56:21 Server kernel: sd 2:0:0:0: [sdb] Mode Sense: 00 3a 00 00

Oct 13 08:56:21 Server kernel: sd 2:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA

Oct 13 08:56:21 Server kernel:  sdb:

 

And then /dev/sdc  (your flash drive)

ct 13 08:56:21 Server kernel: ata4: SATA link down (SStatus 0 SControl 300)

Oct 13 08:56:21 Server kernel: scsi 0:0:0:0: Direct-Access    Kingston DataTraveler 2.0 PMAP PQ: 0 ANSI: 0 CCS

Oct 13 08:56:21 Server kernel: usb-storage: device scan complete

Oct 13 08:56:21 Server kernel: sd 0:0:0:0: [sdc] 3921920 512-byte logical blocks: (2.00 GB/1.87 GiB)

Oct 13 08:56:21 Server kernel: sd 0:0:0:0: [sdc] Write Protect is off

Oct 13 08:56:21 Server kernel: sd 0:0:0:0: [sdc] Mode Sense: 23 00 00 00

Oct 13 08:56:21 Server kernel: sd 0:0:0:0: [sdc] Assuming drive cache: write through

Oct 13 08:56:21 Server kernel: sd 0:0:0:0: [sdc] Assuming drive cache: write through

Oct 13 08:56:21 Server kernel:  sdc:

 

Then, I see the array starting (A parity disk and 1 data disk):

Oct 13 08:56:22 Server emhttp: Device inventory:

Oct 13 08:56:22 Server emhttp: pci-0000:00:08.0-scsi-0:0:0:0 host1 (sda) WDC_WD10EADS-00M2B0_WD-WCAV51053533

Oct 13 08:56:22 Server emhttp: pci-0000:00:08.0-scsi-1:0:0:0 host2 (sdb) WDC_WD10EARS-00Y5B1_WD-WMAV51029159

Oct 13 08:56:22 Server emhttp: shcmd (2): modprobe -rw md-mod 2>&1 | logger

Oct 13 08:56:22 Server emhttp: shcmd (3): modprobe md-mod super=/boot/config/super.dat slots=8,0,0,0,0,0 2>&1 | logger

Oct 13 08:56:22 Server kernel: xor: automatically using best checksumming function: pIII_sse

Oct 13 08:56:22 Server kernel:    pIII_sse  : 11843.200 MB/sec

Oct 13 08:56:22 Server kernel: xor: using function: pIII_sse (11843.200 MB/sec)

Oct 13 08:56:22 Server kernel: md: unRAID driver 0.95.4 installed

Oct 13 08:56:22 Server emhttp: Spinning up all drives...

Oct 13 08:56:22 Server kernel: md: import disk0: [8,0] (sda) WDC WD10EADS-00M WD-WCAV51053533 offset: 63 size: 976762552

Oct 13 08:56:22 Server kernel: md: disk1 removed

Oct 13 08:56:22 Server kernel: mdcmd (2): set md_num_stripes 1280

Oct 13 08:56:22 Server kernel: mdcmd (3): set md_write_limit 768

Oct 13 08:56:22 Server kernel: mdcmd (4): set md_sync_window 288

Oct 13 08:56:22 Server kernel: mdcmd (5): set spinup_group 0 0

Oct 13 08:56:22 Server kernel: mdcmd (6): spinup 0

Oct 13 08:56:22 Server emhttp: shcmd (4): /usr/local/sbin/set_ncq sda 1 >/dev/null

Oct 13 08:56:22 Server kernel: mdcmd (: start STOPPED

Oct 13 08:56:22 Server kernel: unraid: allocating 13240K for 1280 stripes (2 disks)

Oct 13 08:56:22 Server kernel: md1: running, size: 976762552 blocks

Oct 13 08:56:22 Server emhttp: shcmd (5): udevadm settle

Oct 13 08:56:22 Server emhttp: shcmd (6): mkdir /mnt/disk1

Oct 13 08:56:22 Server emhttp: shcmd (7): set -o pipefail ; mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md1 /mnt/disk1 2>&1 | logger

Oct 13 08:56:22 Server kernel: mdcmd (10): check

Oct 13 08:56:22 Server kernel: md: recovery thread woken up ...

Oct 13 08:56:22 Server kernel: REISERFS (device md1): found reiserfs format "3.6" with standard journal

Oct 13 08:56:22 Server kernel: REISERFS (device md1): using ordered data mode

Oct 13 08:56:22 Server kernel: REISERFS (device md1): journal params: device md1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

Oct 13 08:56:22 Server kernel: REISERFS (device md1): checking transaction log (md1)

Oct 13 08:56:22 Server kernel: REISERFS (device md1): Using r5 hash to sort names

Oct 13 08:56:22 Server emhttp: resized: /mnt/disk1

 

I then see transmission starting.... (This might be an issue, since the array disks are not yet mounted)

 

I see unRAID detecting the new disk1 and starting the process of re-constructing it:

Oct 13 08:56:38 Server emhttp: Spinning up all drives...
Oct 13 08:56:38 Server emhttp: shcmd (16): sync
Oct 13 08:56:38 Server kernel: mdcmd (15): spinup 0
Oct 13 08:56:38 Server emhttp: shcmd (17): umount /mnt/user >/dev/null 2>&1
Oct 13 08:56:38 Server emhttp: shcmd (18): rmdir /mnt/user >/dev/null 2>&1
Oct 13 08:56:38 Server emhttp: shcmd (19): umount /mnt/disk1 >/dev/null 2>&1
Oct 13 08:56:38 Server emhttp: shcmd (20): rmdir /mnt/disk1 >/dev/null 2>&1
Oct 13 08:56:38 Server kernel: mdcmd (16): stop 
Oct 13 08:56:38 Server kernel: md1: stopping
Oct 13 08:56:39 Server emhttp: shcmd (21): rm /etc/samba/smb-shares.conf >/dev/null 2>&1
Oct 13 08:56:39 Server emhttp: shcmd (22): cp /etc/exports- /etc/exports
Oct 13 08:56:39 Server emhttp: shcmd (23): /etc/rc.d/rc.samba start | logger
Oct 13 08:56:39 Server logger: Starting Samba:  /usr/sbin/nmbd -D
Oct 13 08:56:39 Server logger:                  /usr/sbin/smbd -D
Oct 13 08:56:39 Server emhttp: shcmd (24): /etc/rc.d/rc.nfsd start | logger
Oct 13 08:56:43 Server emhttp: get_fstype: open /dev/sdb1: No such file or directory
Oct 13 08:56:43 Server emhttp: shcmd (25): modprobe -rw md-mod 2>&1 | logger
Oct 13 08:56:43 Server kernel: md: unRAID driver removed
Oct 13 08:56:43 Server emhttp: shcmd (26): modprobe md-mod super=/boot/config/super.dat slots=8,0,8,16,0,0 2>&1 | logger
Oct 13 08:56:43 Server kernel: xor: automatically using best checksumming function: pIII_sse
Oct 13 08:56:43 Server kernel:    pIII_sse  : 11853.200 MB/sec
Oct 13 08:56:43 Server kernel: xor: using function: pIII_sse (11853.200 MB/sec)
Oct 13 08:56:43 Server kernel: md: unRAID driver 0.95.4 installed
Oct 13 08:56:43 Server kernel: md: import disk0: [8,0] (sda) WDC WD10EADS-00M WD-WCAV51053533 offset: 63 size: 976762552
Oct 13 08:56:43 Server kernel: md: import disk1: [8,16] (sdb) WDC WD10EARS-00Y WD-WMAV51029159 offset: 63 size: 976762552
Oct 13 08:56:43 Server kernel: md: disk1 replaced
Oct 13 08:56:43 Server kernel: mdcmd (2): set md_num_stripes 1280
Oct 13 08:56:43 Server kernel: mdcmd (3): set md_write_limit 768
Oct 13 08:56:43 Server kernel: mdcmd (4): set md_sync_window 288
Oct 13 08:56:43 Server kernel: mdcmd (5): set spinup_group 0 0
Oct 13 08:56:43 Server kernel: mdcmd (6): set spinup_group 1 0
Oct 13 08:56:50 Server emhttp: shcmd (27): /usr/local/sbin/set_ncq sda 1 >/dev/null
Oct 13 08:56:50 Server emhttp: shcmd (28): /usr/local/sbin/set_ncq sdb 1 >/dev/null
Oct 13 08:56:50 Server emhttp: writing mbr on disk 1 (/dev/sdb)
Oct 13 08:56:50 Server emhttp: re-reading /dev/sdb partition table
Oct 13 08:56:50 Server kernel:  sdb: sdb1
Oct 13 08:56:51 Server kernel: mdcmd (10): start RECON_DISK
Oct 13 08:56:51 Server kernel: unraid: allocating 13240K for 1280 stripes (2 disks)
Oct 13 08:56:51 Server kernel: md1: running, size: 976762552 blocks
Oct 13 08:56:51 Server emhttp: shcmd (29): udevadm settle
Oct 13 08:56:51 Server emhttp: shcmd (30): mkdir /mnt/disk1
Oct 13 08:56:51 Server emhttp: shcmd (31): set -o pipefail ; mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md1 /mnt/disk1 2>&1 | logger
Oct 13 08:56:51 Server kernel: mdcmd (13): check 
Oct 13 08:56:51 Server kernel: md: recovery thread woken up ...
Oct 13 08:56:51 Server kernel: md: recovery thread rebuilding disk1 ...

 

I see the server stopping with disk1 missing:

Oct 13 08:57:22 Server kernel: mdcmd (42): stop

Oct 13 08:57:22 Server kernel: md: md_do_sync: got signal, exit...

Oct 13 08:57:22 Server kernel: md: recovery thread sync completion status: -4

Oct 13 08:57:22 Server kernel: md1: stopping

Oct 13 08:57:23 Server emhttp: shcmd (45): rm /etc/samba/smb-shares.conf >/dev/null 2>&1

Oct 13 08:57:23 Server emhttp: shcmd (46): cp /etc/exports- /etc/exports

Oct 13 08:57:23 Server emhttp: shcmd (47): /etc/rc.d/rc.samba start | logger

Oct 13 08:57:23 Server logger: Starting Samba:  /usr/sbin/nmbd -D

Oct 13 08:57:23 Server logger:                  /usr/sbin/smbd -D

Oct 13 08:57:23 Server emhttp: shcmd (48): /etc/rc.d/rc.nfsd start | logger

Oct 13 08:57:31 Server emhttp: shcmd (49): modprobe -rw md-mod 2>&1 | logger

Oct 13 08:57:31 Server kernel: md: unRAID driver removed

Oct 13 08:57:31 Server emhttp: shcmd (50): modprobe md-mod super=/boot/config/super.dat slots=8,0,0,0,0,0 2>&1 | logger

Oct 13 08:57:31 Server kernel: xor: automatically using best checksumming function: pIII_sse

Oct 13 08:57:31 Server kernel:    pIII_sse  : 11853.200 MB/sec

Oct 13 08:57:31 Server kernel: xor: using function: pIII_sse (11853.200 MB/sec)

Oct 13 08:57:31 Server kernel: md: unRAID driver 0.95.4 installed

Oct 13 08:57:31 Server kernel: md: import disk0: [8,0] (sda) WDC WD10EADS-00M WD-WCAV51053533 offset: 63 size: 976762552

Oct 13 08:57:31 Server kernel: md: disk1 missing

 

You start the array once more, but apparently cancel the re-construction of disk1

Oct 13 12:59:27 Server kernel: md: recovery thread rebuilding disk1 ...

Oct 13 12:59:27 Server emhttp: shcmd (78): set -o pipefail ; mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md1 /mnt/disk1 2>&1 | logger

Oct 13 12:59:27 Server kernel: REISERFS (device md1): found reiserfs format "3.6" with standard journal

Oct 13 12:59:27 Server kernel: REISERFS (device md1): using ordered data mode

Oct 13 12:59:27 Server kernel: md: using 1152k window, over a total of 976762552 blocks.

Oct 13 12:59:27 Server kernel: REISERFS (device md1): journal params: device md1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

Oct 13 12:59:27 Server kernel: REISERFS (device md1): checking transaction log (md1)

Oct 13 12:59:27 Server kernel: REISERFS (device md1): Using r5 hash to sort names

Oct 13 12:59:28 Server emhttp: shcmd (79): rm /etc/samba/smb-shares.conf >/dev/null 2>&1

Oct 13 12:59:28 Server emhttp: shcmd (80): cp /etc/exports- /etc/exports

Oct 13 12:59:28 Server emhttp: shcmd (81): mkdir /mnt/user

Oct 13 12:59:28 Server emhttp: shcmd (82): /usr/local/sbin/shfs /mnt/user  -o noatime,big_writes,allow_other,default_permissions

Oct 13 12:59:29 Server emhttp: shcmd (83): killall -HUP smbd

Oct 13 12:59:29 Server emhttp: shcmd (84): /etc/rc.d/rc.nfsd restart | logger

Oct 13 13:00:28 Server emhttp: shcmd (85): /etc/rc.d/rc.samba stop | logger

Oct 13 13:00:28 Server emhttp: shcmd (86): /etc/rc.d/rc.nfsd stop | logger

Oct 13 13:00:29 Server emhttp: Spinning up all drives...

Oct 13 13:00:29 Server emhttp: shcmd (87): sync

Oct 13 13:00:29 Server kernel: mdcmd (34): spinup 0

Oct 13 13:00:29 Server kernel: mdcmd (35): spinup 1

Oct 13 13:00:29 Server emhttp: shcmd (88): umount /mnt/user >/dev/null 2>&1

Oct 13 13:00:29 Server emhttp: shcmd (89): rmdir /mnt/user >/dev/null 2>&1

Oct 13 13:00:29 Server emhttp: shcmd (90): umount /mnt/disk1 >/dev/null 2>&1

Oct 13 13:00:29 Server emhttp: shcmd (91): rmdir /mnt/disk1 >/dev/null 2>&1

Oct 13 13:00:29 Server kernel: mdcmd (37): stop

Oct 13 13:00:29 Server kernel: md: md_do_sync: got signal, exit...

Oct 13 13:00:29 Server kernel: md: recovery thread sync completion status: -4

Oct 13 13:00:29 Server kernel: md1: stopping

 

It is attempting to re-construct disk1, but it is being stopped.  Are you stopping it?

 

Joe L.

I am stopping it as it never asked to format the drive first.  Second I never see the size of the hard drive decreasing as if it is filling.  Previous (yesterday) I let it run it's reconstructing course only to end up with nothing on the disk.

 

Kryspy

I decided to start over as I have my data on spare drives lying around.  This time after a parity check is complete I am going to torture test to make sure I fully understand how this all works and can do a reconstruct.

 

Kryspy

I am stopping it as it never asked to format the drive first.   Second I never see the size of the hard drive decreasing as if it is filling.   Previous (yesterday) I let it run it's reconstructing course only to end up with nothing on the disk.

 

Kryspy

When re-constructing a disk it is NEVER going to ask to format it first.  It is re-constructing the replacement disk with the bit-by-bit image of the disk that failed/was removed.  That image includes everything, including the formatting.    When the re-construction is done, the replacement disk has the structures that would have been there if it had been formatted.

 

I'm afraid you've been shooting yourself in the foot by stopping it.   A re-construct does NOT format a disk and then copy just the files.  It copies every single bit of the old disk to the replacement and that includes the file-system structures that represent the formatting.  A re-construct has absolutely no concept of files... to it, it is just a series of bits.  Every bit of the replacement disk is written, regardless if it was full of files or empty.  You will not see the size changing.

 

When re-constructing a disk, do not stop it.

 

Joe L.

I see.  So is it a one shot deal then for reconstructing data?    I think the problem may have been a strange course of events.    I did this disk update (pin 7/8) on Monday morning and then let it construct.  We went to my mothers for thanksgiving dinner and when we came home the power had gone out somewhere around 4:10 pm which was within 20 minutes of my estimate of the reconstruct being finished. 

 

When I switched the server back on (not attached to a UPS) it began a parity check which I stopped at the 4% mark.  I checked my disk1 and it was blank.  From there I ran a reconstruct again with the same results.   

 

Any chance the power outage affected parity (stupid question) and the reconstruct hadn't actually finished?    I now have a used APC UPS I plan to put into use today.

 

Kryspy

Well,

 

All is working as advertised now.  Re-setup the server.  Ran a parity check.  Removed a disk from the array and re-added it, let it rebuild.  Shares were still acessible while re-building via the virtual disk.

 

Kryspy