-Came home from work to find 2 drives showing unformatted

-ran reisfsck –check to find it say hardware problem with unreadable sectors

-reisfsck recommends replacing drives due to hardware problem

-Drive 6 and 9 originally could be accessed with missing files, but now totally unaccessable

-I have a Hatachi 2TB in closet and WD 2TB on order to replace the bad drives

-Was reading the forums before doing anything and came across posts about problems with 2TB Drives

-I’m still reading the threads on large drive issues and realized I need help to not make the problem worst.

 

So my main question is how best should I proceed to replace these two damaged drives (disk 6 and disk 9) in my system?

 

A long term question is how best to deal with the 2TB drives already in my system working just fine?

 

Thanks in Advance.  Any and all help is greatly appreciated.

 

I don’t read the forums as much as I should, mainly because my system runs so smoothly.  I had no idea there was an issue with some larger Hard Drives until now that I am having an issue.

 

I came home from work Friday to find my server unresponsive.  I powered down and restarted to find Drives 6 and 9 showing as unformatted.  I used putty to telnet into my system and run reisfsck –check command which returned information that both drives had possible hardware problems with unreadable sectors.

Disk 9 has a red dot by it and Disk 6 has a green dot by it on the disk status page.  Now neither drive is loading when I restart the array.  On unmenu Disk 9 has DISK_DSBL_NP and Disk 6 has DISK_NP_MISSING.  Both disks are Samsung HD154UI 1.5 TB drives purchased and installed about 1 year ago.  When I reboot the array will not start as it shows both drives as missing.

 

I didn’t have a power outage as my main computer was still up and running when I got home and my server is on a UPS.  I do find it unusual for 2 of the exact same model hard drives to fail at the same time, but for now that is what the data is showing.  I have opened up the system and reseated all my drives and sata connections and reseated the memory and expansion cards.  I also took the flash drive and scanned it for errors and found none.

 

I am looking to buy one Western Digital 2TB (WE20EARS) drive and I already have one HITACHI HDS722020ALA330 / 0F10311 Deskstar 7K2000 2TB which I plan to use to to replace the 2 bad identified drives.  I am running low on free disk space at this time, but I do have room to move the contents of the 2 bad drives to safe locations.  However, at this time most of the files are not showing up on either drive to move.

 

I will also include 2 syslogs.  The first log was early in the problem when I was able to boot and still access some part of the bad 2 drives.  The 2nd syslog is most current and shows that I can’t boot at all as both bad drives show as missing.

 

Since I just became aware of an issue with some larger hard drives, I have not had a chance to read all the threads.  I’m hoping for some suggestions on how best to proceed to avoid problems in the future.  Below I have provided some information about my system and the drives:

 

 

Supermicro MBD-C2SEE-O  775 Intel G43 Motherboard

4 Gig OCZ Memory

Intel E5200 Wolfdale 2.5GHz Dual Core Processor

1 Thermalake 550W Powersupply

1Thermalake 650W Powersupply

 

Parity – Hitachi HDS 72202 2TB

Disk 1 – Hitachi HDS 721075KLA330

Disk 2 – Hitachi HDS 721075KLA330

Disk 3 – Hitachi HDS 72202  2TB

Disk 4 – Hitachi HDS 721075KLA330  

Disk 5 – Hitachi HDS 721075KLA330

Disk 6 --  Samsung HD154UI  

Disk 7 – Samsung HD 103SI

Disk 8 – Seagate ST32000542AS 2TB

Disk 9 --  Samsung HD154UI

Disk 10 – Samsung HD753Lj

Disk 11 – Samsung HD 103SI

 

syslog-2011-01-22_last.txt

Be careful -- I had almost identical symptoms to yours and after replacing the SATA and power cables, I finally worked out that two of my SATA ports had died.

 

I put the "failed" drives on another motherboard and found all the data intact.

There was only one syslog attached, so I could not compare the earlier one with the later, but I don't think it will matter.  This syslog appeared to be the later one, because both drives are not found.  The important indicator here may be good news, in that what actually is missing is the disk controller that the 2 drives shared!  In other words, the 2 drives may be fine, just a failed disk controller card.  All 6 drives on the motherboard are seen and loaded, and all 4 drives on the 4 port SATA card are loaded, but there is an indication of another disk controller with at least 2 SATA ports that is not operational, and of course, that makes the attached drives invisible.

 

I have to assume for now that reiserfsck was confused, because of the malfunctioning card, when it tried to blame 'unreadable sectors'.  The sectors may have been just unavailable.

 

Because this involves 2 drives, I would proceed with extreme caution.  I personally would not even turn this array on, until you can locate and install another disk controller.  I am very hopeful though that the 2 drives will reappear without any serious issues, once connected to a working disk controller.  Let us know ...

Just an update...I do think you are correct regarding a dead controller card.  I ordered a replacement card off newegg with overnight delivery to arrive Tuesday.  I live in a very small town that is proud of its 3 Walmarts.  I spent the day looking for a replacement card with no luck.  It was almost funny watching the look on the sales rep face when I asked if they carried an internal SATA II Controller card...they would walk around to the printers...then the video cards...before asking what the card is used for.

 

Update on Tuesday.  Thanks again for the help.

I like Walmart for some things, but the computer section is definitely not one of them!  Selection is poor, offered product choices are bad, quality is poor, generally non-existent experienced help, and even the prices are terrible, which is especially hard to understand.  For store fronts, I like CompUSA best, then SamsClub or CostCo, then Staples and Office Depot and Office Max somewhat behind, and Best Buy far to the rear (how the mighty have fallen).  SamsClub's choices are very limited, but generally good, although often overpriced compared to Newegg.  Of course, if you have a little time, then nothing beats Newegg and Geeks.com and others.

Monoprice also has a 2 port PCIe x1 card reported to work well.

--Still showing a problem with Disk 9

--Shows Red Dot by name and on unmenu shows DISK_DSBL

--Disk 6 now shows to be working fine

--I can now access BOTH disks and all files seem to be there

--Before I do anything else, I'm moving the files off disk 9 to a different disk

--a new syslog is included here

 

I got the card in from Newegg and installed and immediately it showed both disk again and the array booted up.  Disk 6 shows all green and shows no problems now.  Disk 9 still has the red dot and shows DISK_DSBL on unmenu.  I have not run Reisfsdk because before doing anything else I am moving the files from disk 9 since its now readable.

 

Tomorrow I plan to go replace the sata cable on disk 9 and move that drive to one of the open sata ports on the new card I just got from newegg and see if that fixes the problem.

 

 

Thanks again for the help.  No matter what else I find, I'm at least back to the point of being able to save my data before moving forward.  Once this problem is fixed, I will move to the new version of unraid since I am still running four 2 TB drives which so far seem to be working with no issues.  I will update tomorrow once I open up my system.  It will take all nite and most of tomorrow to finish moving files off Disk 9.

syslog-2011-01-25.txt

Something about your hardware configuration makes me wonder if there is another reason.

 

1.You have a Supermicro motherboard (traditionally considered to be very picky in regards to the memory used) and OCZ brand DDR3 (which I personally consider one of the worst to get - but you do not have to agree with me on that one ) Thanks god they are out of the DRAM business now.

 

Supermicro's are considered "server" grade and they are after the stability and I believe they have very limited if any options to accommodate any relaxing from the timings and the standard voltages.

 

OCZ is exactly opposite as they are after the "enthusiastic" end of the market where the people are overclocking madly (but with not much regards for long therm stability). They will buy the cheapest ICs, assemble them into modules and then will program, test and sale these modules in tens if not hundreds variations (and often they require custom timing and voltage settings that a Supermicro board does not have)

 

You said you did not have any problems in the past so perhaps you were lucky and got a good set of memory but this is something to keep in mind and if never done please run the memtest.

 

2. Why you have two PSU - did you use originally the smaller one and then switched to the 650W one or you use some kind of power sharing???

Without the exact model number it is a hard to guess if you should look into possible problems here.

 

3. I will add here - disable the onboard IDE port in the BIOS

 

--Still showing a problem with Disk 9

--Shows Red Dot by name and on unmenu shows DISK_DSBL

--Disk 6 now shows to be working fine

--I can now access BOTH disks and all files seem to be there

--Before I do anything else, I'm moving the files off disk 9 to a different disk

--a new syslog is included here

 

That is quite normal, unRAID does not know what you have been doing, so it still has the disk marked as disabled in its own config.  You have just one more step and the array should be back up fine.  Just do the Trust My Array procedure, and let the parity check complete.  This procedure lets you inform unRAID that the disks are actually fine.

 

The new syslog looks good, no problems noted.

Sorry for the delay in responding...I got sent out of town and am just now getting back.

 

--I replaced the SATA Cable

--Trusted the Array

--Everything is showing green and working fine

--Going to rebuild parity next, and upgrade to new version of Unraid as well as install a new 2T WD drive

 

Thanks again for all the help.  Your comments and suggestions were dead on correct.  The array is now back up and running showing all green.  It's like having one of your kids get sick and watching them slowly get well.

 

Next thing I plan to do (after another out of town trip) is rebuild parity and then install that extra 2TB WD EARS drive I bought in case drive 9 was bad.  I have heard about the issues with the larger drives and I currently have Three (3) larger 2TB drives in my system now including the parity drive.  They have been in place and working fine for over 8 months now (I didn't mess with the jumpers on the drives).  

 

I plan on updating to the new version of unraid and then install the last 2TB drive I have.  I had not planned on messing with the existing 3 larger drives that seem to be working fine.  If I"m doing this wrong please let me know.  Otherwise this weekend I will do the upgrade when I get home.

 

Thanks again for all the help getting me back up and running with no loss.  It's kinda amazing how much you all were able to divine and diagnose from those syslogs.  Have a good one and for all you up north...stay warm.