April 4, 201115 yr Hi, just wan't a quick confirmation, im running 4.5.6 right now and one of my old Samsung HD203WI Drives seems to making problems (write errors and marked as red). Due to the fact that i can't buy any new ones of this modell I thought of switching to the 4k 204er Modell, all others are non-4k-drives. I would like the most hassle free approach here, can I update unraid to 4.7 and rebuild after onto the new 204er Samsung or is this not recommended? Thanks...
April 4, 201115 yr This sounds like you've got a problem with your drive. Upgrading to 4.7 won't help with that... it will only complicate matters! Please post a SMART report on the drive, and syslogs with relevant events.
April 4, 201115 yr I would like the most hassle free approach here, can I update unraid to 4.7 and rebuild after onto the new 204er Samsung or is this not recommended? You could upgrade first, but as mentioned, that complicates it a tiny bit as you will not have an easy way to regress to the prior versions of unRAID once you format a drive to 4k-alignment. Best would be to replace the drive in the current version of unRAID (assuming it really is bad and it is not just a loose cable) and then upgrade after parity protection has been re-established. After you upgrade you can, if you feel confident in the array, un-assign the new drive, change the partitioning, and then re-assign it and let unRAID re-construct the contents based on parity and all the other data disks. Joe L.
April 4, 201115 yr Author Thanks for the suggestions... @christuf I know, this wouldn't solve the problem... but I thought its easier without fighting the 4k topic on an old installation. I'll attached a syslog and a smart report for the device... i cut out 10MB of repeating errors like this Apr 3 10:33:49 SPACEHOG kernel: ata11.00: device reported invalid CHS sector 0 Apr 3 10:33:49 SPACEHOG kernel: ata11: status=0x41 { DriveReady Error } Apr 3 10:33:49 SPACEHOG kernel: ata11: error=0x04 { DriveStatusError } Apr 3 10:33:49 SPACEHOG kernel: ata11: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 bevor it shows read errors after that the drive wasn't accessible at all, maybe the system shut it down? I don't know....but the smart report don't indicate anything, atleast how far I can interpret it... unraid.zip
April 4, 201115 yr Thanks for the suggestions... @christuf I know, this wouldn't solve the problem... but I thought its easier without fighting the 4k topic on an old installation. I'll attached a syslog and a smart report for the device... i cut out 10MB of repeating errors like this Apr 3 10:33:49 SPACEHOG kernel: ata11.00: device reported invalid CHS sector 0 Apr 3 10:33:49 SPACEHOG kernel: ata11: status=0x41 { DriveReady Error } Apr 3 10:33:49 SPACEHOG kernel: ata11: error=0x04 { DriveStatusError } Apr 3 10:33:49 SPACEHOG kernel: ata11: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 bevor it shows read errors after that the drive wasn't accessible at all, maybe the system shut it down? I don't know....but the smart report don't indicate anything, atleast how far I can interpret it... have you power cycled since the failure? We've seen some disks stop responding, but wake again after a power cycle. Joe L.
April 4, 201115 yr The problems are not with the Samsung; It is the md5 = ata11 = sdl = WDC WD20EADS-00S WD-WCAVY3682149 Added - You have pretty loaded configuration. You can power down and check the connections (especially the power splitters). It is no harm to check tha data cable too - IMHO the brand name "forward breakout" cables use regular non-locking SATA connectors. Another editing - you are on 4.5.4
April 4, 201115 yr Author @bcbgboy13 sorry, mixed this up while posting (where at work) @Joe L. yeah the disk is back up, but of course still red...but since i copied data to it while it might was red, a check and trust my array procedure wouln't be out of the question (as far as i understand it it would result in data loss if I ran a parity check again). Is there maybe a procedure to preclear the disc and rebuild it again?
April 4, 201115 yr You're welcome - I must have misunderstood your initial question. When I read it, it sounded like you thought a v.4.7 upgrade would solve the issue you were having with your drive. From the SMART report, that drive looks ok (no sectors reallocated or pending reallocation), but it's not a Samsung... I really don't know what I'm doing with syslog yet, so will let someone with more experience pass comment on that front. With many many years of IT experience, I have learned to err on the side of caution over speed. Yes, it would be faster to upgrade to 4.7 and sort out your disk issue and 4k alignment at the same time, but you risk hitting a problem and data loss. Much in the same way I would love to just put my new 2TB drive into the array to play with, rather than waiting 20-30 hours for a preclear!
April 4, 201115 yr Is there maybe a procedure to preclear the disc and rebuild it again? Yes, you stop the array, unassign the disk, start the array, stop the array, assign the disk and finally start the array again. The drive will then be rebuilt. Peter
April 4, 201115 yr Author @christuf my intention with the upgrade to 4.7 was not to worry about jumpers and allignements with further upgrades to new 4k discs. yeah i mixed up the brands since I where at work when I posted. The connections are hopefully fine, I rebuilded the server with most efficient way of cabeling in mind half a year ago to prevent these kind of errors. @lionelhutz sounds good this would be my prefered next step
April 5, 201115 yr Author I rebuild the array on the WD again, and everythings looking fine... strange. Anyway thanks a lot for your support all! Going to upgrade now to 4.7... to be ready for further hdd crashs.
Archived
This topic is now archived and is closed to further replies.