Ok so i installed 2 new drives a couple days ago, making one of these new drives the parity drive. But i didn't assign the parity drive straight away as i was going to be doing alot of moving of files to clean things up first, which i did over the last couple of days.

 

Last night i assigned the parity drive and started a parity sync.. it's all finished and there are 24 errors reported for one of the drives.

 

Now one of the iso's that i tried to moved wouldn't it was failing 1/3 of the way through i tried a few times in mc and also from within windows onto main computer, so i ended up just deleting the file wasn't that important.

 

I'm now guessing this drive must have some bad sectors on it???

 

I've attached the syslog if that tells anyone in the know what the errors were caused by.

 

 

Disk error ? I think so when i se those lines :

May 30 03:28:43 Server kernel: ata9.00: configured for UDMA/133
May 30 03:28:43 Server kernel: ata9: EH complete
May 30 03:28:44 Server kernel: ata9.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
May 30 03:28:44 Server kernel: ata9.00: BMDMA stat 0x24
May 30 03:28:44 Server kernel: ata9.00: cmd 25/00:48:c7:e7:c2/00:02:25:00:00/e0 tag 0 dma 299008 in
May 30 03:28:44 Server kernel:          res 51/40:00:53:e9:c2/40:00:25:00:00/e0 Emask 0x9 (media error)
May 30 03:28:44 Server kernel: ata9.00: status: { DRDY ERR }
May 30 03:28:44 Server kernel: ata9.00: error: { UNC }
May 30 03:28:45 Server kernel: ata9.00: configured for UDMA/133
May 30 03:28:45 Server kernel: ata9: EH complete
May 30 03:28:46 Server kernel: ata9.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
May 30 03:28:46 Server kernel: ata9.00: BMDMA stat 0x24
May 30 03:28:46 Server kernel: ata9.00: cmd 25/00:48:c7:e7:c2/00:02:25:00:00/e0 tag 0 dma 299008 in
May 30 03:28:46 Server kernel:          res 51/40:00:53:e9:c2/40:00:25:00:00/e0 Emask 0x9 (media error)
May 30 03:28:46 Server kernel: ata9.00: status: { DRDY ERR }

May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530640/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530648/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530656/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530664/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530672/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530680/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530688/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530696/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530704/5, count: 1
May 30 03:28:52 Server kernel: md: disk5 read error
May 30 03:28:52 Server kernel: handle_stripe read error: 633530712/5, count: 1

 

thoses errors are reccurence.

He's right.  The key words there are media error and UNC (uncorrectable).  It appears there is a small cluster of bad sectors, which you hit 4 times, three times before the parity build, and once during.  I would run the SMART long test on Disk 5, some instructions are at the bottom of the Obtaining a SMART report section.  For large drives, it can take a long time.

thanks guys, i pulled all the data off the drive last night and then ran a full surface scan with samsungs diag tool. when i got up this morning it said it found errors and said to do a full erase, which should be done by now. so when i get home i'll might try booting back up into unraid (so i can watch some stargate ) and let that smart long test run see what it reports then... does that test do a full surface scan as well?.. or should i just run the samsung one again to see if it's fixed the errors?

i'm about to head home in 10 and was checking that smart test info.

 

The part of the command smartctl -d ata -tlong /dev/sda... i beleive from memory of other threads i've read, it needs to be changed to corrispond to my disk 5 one correct? if so how do i find that info out of what i specificly need to put in to make sure i scan the correct disk.

i'm about to head home in 10 and was checking that smart test info.

 

The part of the command smartctl -d ata -tlong /dev/sda... i beleive from memory of other threads i've read, it needs to be changed to corrispond to my disk 5 one correct? if so how do i find that info out of what i specificly need to put in to make sure i scan the correct disk.

Go to your "Devices" page in the management interface.  The three letters in the parens are the correct device.

 

So if you see something like this:

disk5 device:  pci-0000:00:1f.2-scsi-0:0:0:0 (sdf) ata-WDC_WD10EACS-00D6B1_WD-WCAU44206983

the device is /dev/sdf

 

Joe L.

ok i think it's running, guess i'll wait a few hours and see what happens 

 

well that smart scan thing did nothing.. i'm sure i did it right, it told me to wait 190 something mins for the scan but the disc spun down and nothing ever happened let it go for several hours to i went to bed, and nothing... so i shut down and started samsungs utility back up and let that do it's thing and all seems to be ok now acording to that, so only thing left i spose to do is dump all the files back onto it and do another parity and see what happens.. fingers crossed

well that smart scan thing did nothing.. i'm sure i did it right, it told me to wait 190 something mins for the scan but the disc spun down and nothing ever happened let it go for several hours to i went to bed, and nothing... so i shut down and started samsungs utility back up and let that do it's thing and all seems to be ok now acording to that, so only thing left i spose to do is dump all the files back onto it and do another parity and see what happens.. fingers crossed

 

The long test does not automatically output its results.  You have to tell it to.  I'm sure it is in the wiki somewhere... or a quick google search will tell you how to get at the results.

 

Sorry, it's late and I need to get to bed so I can get up for class tomorrow

well parity finished and no errors were reported.. so hopefully i'm all in the clear, whatever the problem was it must of just been superficial, and the low level format fixed up..