Pre-aligned advanced format drive is giving errors w/jumper on boot.

The WD20EARS advanced format drive that had been aligned on Windows XP using the WD Drive Alignment Tool is giving me errors when I applied a jumper to pin 7 & 8.  It is fine without the jumper.

So is it possible that the WD Align Tool is permanent and therefore no jumper is needed?  Can someone hint to as what the errrors are about?

Here are a sampling of the errors:

Oct 16 15:35:25 Beanstalk kernel: ata20.00: exception Emask 0x0 SAct 0x2 SErr 0x0 action 0x0

Oct 16 15:35:25 Beanstalk kernel: ata20.00: irq_stat 0x00020002, device error via SDB FIS

Oct 16 15:35:25 Beanstalk kernel: ata20.00: failed command: READ FPDMA QUEUED

Oct 16 15:35:25 Beanstalk kernel: ata20.00: cmd 60/08:08:a8:88:e0/00:00:e8:00:00/40 tag 1 ncq 4096 in

Oct 16 15:35:25 Beanstalk kernel:          res 41/01:00:af:88:e0/ff:00:e8:00:00/40 Emask 0x401 (device error) <F>

Oct 16 15:35:25 Beanstalk kernel: ata20.00: status: { DRDY ERR }

Oct 16 15:35:25 Beanstalk kernel: ata20.00: configured for UDMA/100

Oct 16 15:35:25 Beanstalk kernel: ata20: EH complete

Oct 16 15:35:25 Beanstalk kernel: ata20.00: exception Emask 0x0 SAct 0x1 SErr 0x0 action 0x0

Oct 16 15:35:25 Beanstalk kernel: ata20.00: irq_stat 0x00060002, device error via SDB FIS

Oct 16 15:35:25 Beanstalk kernel: ata20.00: failed command: READ FPDMA QUEUED

Oct 16 15:35:25 Beanstalk kernel: ata20.00: cmd 60/08:00:a8:88:e0/00:00:e8:00:00/40 tag 0 ncq 4096 in

Oct 16 15:35:25 Beanstalk kernel:          res 41/01:00:af:88:e0/ff:00:e8:00:00/40 Emask 0x401 (device error) <F>

Oct 16 15:35:25 Beanstalk kernel: ata20.00: status: { DRDY ERR }

Oct 16 15:35:25 Beanstalk kernel: ata20.00: configured for UDMA/100

Oct 16 15:35:25 Beanstalk kernel: ata20: EH complete

Oct 16 15:35:25 Beanstalk kernel: ata20.00: exception Emask 0x0 SAct 0x1 SErr 0x0 action 0x0

Oct 16 15:35:25 Beanstalk kernel: ata20.00: irq_stat 0x00060002, device error via SDB FIS

Oct 16 15:35:25 Beanstalk kernel: ata20.00: failed command: READ FPDMA QUEUED

Oct 16 15:35:25 Beanstalk kernel: ata20.00: cmd 60/08:00:a8:88:e0/00:00:e8:00:00/40 tag 0 ncq 4096 in

Oct 16 15:35:25 Beanstalk kernel:          res 41/01:00:af:88:e0/ff:00:e8:00:00/40 Emask 0x401 (device error) <F>

Oct 16 15:35:25 Beanstalk kernel: ata20.00: status: { DRDY ERR }

Oct 16 15:35:25 Beanstalk kernel: ata20.00: configured for UDMA/100

Oct 16 15:35:25 Beanstalk kernel: ata20: EH complete

Oct 16 15:35:25 Beanstalk kernel: ata20.00: NCQ disabled due to excessive errors

Oct 16 15:35:25 Beanstalk kernel: ata20.00: exception Emask 0x0 SAct 0x1 SErr 0x0 action 0x0

Oct 16 15:35:25 Beanstalk kernel: ata20.00: irq_stat 0x00060002, device error via SDB FIS

Oct 16 15:35:25 Beanstalk kernel: ata20.00: failed command: READ FPDMA QUEUED

Oct 16 15:35:25 Beanstalk kernel: ata20.00: cmd 60/08:00:a8:88:e0/00:00:e8:00:00/40 tag 0 ncq 4096 in

Oct 16 15:35:25 Beanstalk kernel:          res 41/01:00:af:88:e0/ff:00:e8:00:00/40 Emask 0x401 (device error) <F>

Oct 16 15:35:25 Beanstalk kernel: ata20.00: status: { DRDY ERR }

Oct 16 15:35:25 Beanstalk kernel: ata20.00: configured for UDMA/100

Oct 16 15:35:25 Beanstalk kernel: ata20: EH complete

Oct 16 15:35:25 Beanstalk kernel: ata20.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0

Oct 16 15:35:25 Beanstalk kernel: ata20.00: irq_stat 0x00060002, device error via D2H FIS

Oct 16 15:35:25 Beanstalk kernel: ata20.00: failed command: READ DMA EXT

Oct 16 15:35:25 Beanstalk kernel: ata20.00: cmd 25/00:08:a8:88:e0/00:00:e8:00:00/e0 tag 0 dma 4096 in

Oct 16 15:35:25 Beanstalk kernel:          res 51/01:00:af:88:e0/ff:00:e8:00:00/e0 Emask 0x1 (device error)

Oct 16 15:35:25 Beanstalk kernel: ata20.00: status: { DRDY ERR }

Oct 16 15:35:25 Beanstalk kernel: ata20.00: configured for UDMA/100

Oct 16 15:35:25 Beanstalk kernel: ata20: EH complete

Oct 16 15:35:25 Beanstalk kernel: ata20.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0

Oct 16 15:35:25 Beanstalk kernel: ata20.00: irq_stat 0x00060002, device error via D2H FIS

Oct 16 15:35:25 Beanstalk kernel: ata20.00: failed command: READ DMA EXT

Oct 16 15:35:25 Beanstalk kernel: ata20.00: cmd 25/00:08:a8:88:e0/00:00:e8:00:00/e0 tag 0 dma 4096 in

Oct 16 15:35:25 Beanstalk kernel:          res 51/01:00:af:88:e0/ff:00:e8:00:00/e0 Emask 0x1 (device error)

Oct 16 15:35:25 Beanstalk kernel: ata20.00: status: { DRDY ERR }

Oct 16 15:35:25 Beanstalk kernel: ata20.00: configured for UDMA/100

Oct 16 15:35:25 Beanstalk kernel: sd 20:0:0:0: [sdj] Unhandled sense code

Oct 16 15:35:25 Beanstalk kernel: sd 20:0:0:0: [sdj] Result: hostbyte=0x00 driverbyte=0x08

Oct 16 15:35:25 Beanstalk kernel: sd 20:0:0:0: [sdj] Sense Key : 0x3 [current] [descriptor]

Oct 16 15:35:25 Beanstalk kernel: Descriptor sense data with sense descriptors (in hex):

Oct 16 15:35:25 Beanstalk kernel:         72 03 13 00 00 00 00 0c 00 0a 80 00 00 00 00 00

Oct 16 15:35:25 Beanstalk kernel:         e8 e0 88 af

Oct 16 15:35:25 Beanstalk kernel: sd 20:0:0:0: [sdj] ASC=0x13 ASCQ=0x0

Oct 16 15:35:25 Beanstalk kernel: sd 20:0:0:0: [sdj] CDB: cdb[0]=0x28: 28 00 e8 e0 88 a8 00 00 08 00

Oct 16 15:35:25 Beanstalk kernel: end_request: I/O error, dev sdj, sector 3907029167

Is there some test I can perform on the disk without the jumper to see if it is properly aligned in unRAID?

why not just run the preclear script from your servers console instead of that  wd align thing on windows.

-=Jason=-

why not just run the preclear script from your servers console instead of that  wd align thing on windows.

 

-=Jason=-

I am migrating the drive from a Windows XP machine to my newly built unRAID server.  The drive had previously been pre-aligned for Windows XP using the WD Drive Alignment Tool.  My intention was to install the jumper on pin 7 & 8 and then run the preclear script.  However, with the jumper applied, unRAID gives me tons of errors on boot and the preclear script just hangs.  I have two other WD20EARS that works fine with the jumper, but they were never aligned with the WD Drive Alignment Tool.

Here is an interesting read on using the parted tool to align these advanced format drives.

Is this possible with unRAID?  If so, it could also solve the Samsung Spinpoint F4 drives.

Fyi i have a few EARS drives that were formatted without the jumper. I reied everything to reformat with the jumper and all i got was simialr errors.

I gave up

My understanding from what I have read is that installing the jumper remaps ALL sectors (granted they are simulated 512 byte logical sectors) to Logical Block Address x - 1. So real sector 0 becomes -1 (can't be addressed), 1 becomes 0, 64 becomes 63 (convienent). This makes LBAs (after translation) 7, 15, ... 63 line up to the physical 4k sector boundries which results in the best performance for partitons that start on one of those boundries.

If a drive is formatted in any system without the jumper and then the jumper is installed everything is off by one 512 byte sector. Drive signatures, partition table, partitions, partition boot sectors, EVERYTHIN. This seems to cause some systems to see that something is on the drive but can't resolve the structure, thus resulting in many errors. On the other had if the drive is completely empty (with or without the jumper installed) systems don't get confused because they see an empty drive.

I have read that people have been able to add the jumper to a prevously formatted drive by using any utiltiy that will completely earase the drive to zeros (preclear does not exactly do this) without the jumper installed and then install the jumper and format or preclear it. Maybe all the problems caused by the jumper including RMAs is why Samsung F4 drives don't have one.

In you case the partition might start on LBA 63 after installing jumber but the rest of the drive structure is still hosed.

Ahh, so because my drive was pre-formatted without the jumpers.  I assumed the WD Disk Diagnostic zero write would have done it, but I guess it does not.  Would fdisk or parted in Linux allow me to completely zero out the drive so that I can intall the jumper?

I have Ubuntu 10.10 that I can boot from a USB flash drive.

I assumed the WD Disk Diagnostic zero write would have done it, but I guess it does not.

That is what I would have tried, without the jumper installed. Then installed the jumper and precleard or formatted it. Sorry, I didn't read where you had already tried that. I would use the CD ISO version not the windows version. No real reason other than trying to to take out as many variables as possible. If that doesn't work there may be more going on internally in the drive than what on the surface seems like a simple sector remap.

There is a lot about the WD advanced format drives in this thread.

http://lime-technology.com/forum/index.php?topic=5384.0

I had similar problems with my WD EAR drive. I had it installed in a Windows 7 machine without a jumper. First I just clean formated the drive and installed into my UnRaid server and got similar errors as you when I installed the jumper.

So without the jumper on I used my windows machine to kill all the partitions on the drive, so it was bare. Then, put the jumper on and installed in my UnRaid server and it booted up without errors. Currently its busy rebuilding parity.

In my case, the WD Align Tool was used on the drive since I run Windows XP.  WD states that a jumper can no longer be used once the drive has been formatted.  I even tried writing zeroes to the drive and still I could not use the jumper.  The only thing that I did not try was to zero the drive with the jumper installed.

I decided to not deal with it and just requested an RMA on the drive.  I do have a WD15EARS that is in the same situation, but I think I will keep it on my Windows XP machine for now instead of migrating it to unRAID.

After returning the first drive where I could not use the jumper on pin 7-8 because the drive had previously been formatted with the jumper, I decided to test the second drive that had also been formatted without the jumper and aligned with the WD Alignment Tool.

This time I decided to install the jumper on pin 7-8 prior to zeroing the drive with the WD Windows Diagnostic utility.  It worked!  Not only did the drive pass the short SMART test with the jumper (the first drive did not) it also did not cause any errors with unRAID.  So it appears you would have to zero the drive with the jumper on.

Edit:

I spoke too soon.  Preclearing the disk is yielding tons of the same error related to installing a jumper on pin 7-8.  Time to RMA this drive too.

I'm having the same exact errors when I put the pin on a drive that has been working in my system for months. I don't get any parity errors, but want my system to be setup correctly.

Right now, I can't get to the unRaid Control Page when the pin is on. I run headless and after hooking up a monitor I was seeing those type of errors. It eventually gets to a logon prompt.

I will try zeroing out the drive. Like other have suggested first and report back.

I'm wondering how you are getting on with this issue as it is potentially a blocker for me to migrate my storage to unRAID.

I have all my media shared out from a XP Pro 32-bit, it works very well but has limits for expansion etc.

I recently purchased 2 x WD15EARS to add to the XP media server, I did the following.

1) I ran the WD Align on the drive without the jumper in place

2) Ran WD short and long SMART test, both drives passed

3) Currently is use in XP machine - both drives can be wiped if needed (a must seeing as unRAID uses ReiserFS)

So, I'm now thinking I would like to use unRAID (it will be on the same hardware platform), I have booted the machine from the free version unRAID to test hardware compatability and all seems fine.

Now I'm stuck, how do I now integrate my 2 x WD15EARS WD Aligned drives?

Maybe using Darik's boot and Nuke on them, then starting over?

Try various tools from the Ultimate Boot CD?

Zero the drive using WDAlign with the jumper set?

Zero the drive using WDAlign without the jumper?

Zero the drive using unRAID with the jumper set?

Zero the drive using unRAID without the jumper set?

Is pre-clear he same as zeroing the drive?

Is pre-clear he same as zeroing the drive?

The preclear_disk.sh script

• 1. reads the entire disk both linearly interspersed with random sectors to exercise the head positioning mechanism's ability to read all sectors and identify un-readable sectors.
  
• 2. zeros the entire disk  (this is the part that is similar)  This should allow the smart firmware on the disk to re-allocate un-readable sectors detected in step 1.  If they had not been identified in step 1, they would not be identified until after you put your precious data in them  :'(
  
• 3. creates an initial disk partition and writes a special "signature" to the master boot record of the disk that allows unRAID to identify the fact it was pre-cleard so it will skip its own clearing step when adding the drive to a protected array.
  
• 4. reads the entire disk confirming that the signature and all the zeros were written correctly.

I can't help you with the jumpers and the drives themselves... I have no experience to draw upon.  I've stayed clear of those drives using Seagate and Hitachi drives instead.

Thanks for the info.

I'm thinking that I might have to run pre-clear without the jumper, so unRAID can 'see' all the sectors. Then run it again with the jumper in place, as using the jumper seems to offer the most compatibility...I think  

Time to do some backing-up and the some testing I guess, I'll let you know the outcome.

My second replacement drive arrived yesterday.  I immediately installed a jumper on pin 7 & 8 and started preclear on the drive.

The best advice I can give for these WD advanced format drives (EAR model) are:

01. Immediately install the jumper on pin 7-8 for Windows XP and unRAID.  Do not bother with the WD Drive Alignment utility.

02. Once the drive is formatted without the jumper, it will be very difficult to get them to work error free in unRAID with the jumper.

In a couple of weeks I will have finished building my unRAID server. Initially it will have 4 x 2TB.

When the server is running fine, I will copy all my data to it, which currently is stored on a Synology DS410. After finishing that, I would like to add the 'old' DS410 drives (4 x WD20EARS without jumper) to my unRAID serve to have all in all 8 x 2TB.

How do I prepare the 'old' WD20EARS for that purpose? When I purchased those drives, I put them directly into the DS410 without any additional preparations. So far, they work fine without any errors.

Maybe it's easier to just sell them with the DS410?

Chris

Milfronte - thanks for the tip, difficult to know what to do, especially when the advice to use them with XP is to use the WDAlign as it gives the best performance

I purchased my drive from Newegg, it is most likely I can still return them, I think. Did you have any issue returning them for replacement?

Milfronte - thanks for the tip, difficult to know what to do, especially when the advice to use them with XP is to use the WDAlign as it gives the best performance

 

I purchased my drive from Newegg, it is most likely I can still return them, I think. Did you have any issue returning them for replacement?

Actually, the instructions on the WD EAR drives say to either use the jumper or the WD Align Utility for Windows XP.  Of course it also says that for other OS'es the drive is ready to go.  I would now use just the jumper since it offers the best compatibility if you want to go from Windows XP to unRAID.  I just don't know if you can apply the jumper for unRAID after  you have used the drive in Windows 7.

I used WD's Warranty replacement.  Selecting the advance replacement option, WD will ship the replacement drive to you before receiving your drive (it took 3 days for me).  You do have to use a credit card as a guarantee, but they don't charge anything to the card unless you fail to return your drive in 30 days.  Going this route vs. a NewEgg return has three benefits:

01. WD package the replacement drive much better than NewEgg.

02. The replacement drive was probably tested prior to being shipped.

03. You are beyond NewEgg's return period (my case).

Your old drive's warranty carries over to the new drive once WD receives your old drive.  I have to say, I am quite impressed with WD warranty replacement process.  Unlike Seagate's process, where they charge you a $20 processing fee for the advance feature, so you are left waiting and waiting and waiting for the replacements drive.

Thanks again.

I use the disks for data only on XP, so given the choice of 2 offered by WD, I chose the WDalign option as there were a few benchmarks out there that indicated that jumpering the drive can have a negative effect on performance, so I guess I made the wrong choice....

Were your original drives faulty? Or just screwed up because of the alignment issue?

Thanks again.

 

I use the disks for data only on XP, so given the choice of 2 offered by WD, I chose the WDalign option as there were a few benchmarks out there that indicated that jumpering the drive can have a negative effect on performance, so I guess I made the wrong choice....

 

Were your original drives faulty? Or just screwed up because of the alignment issue?

My original two drives worked fine under XP (data only).  I too used the WD Align option since I did not own any jumpers at that time.  It was when I applied the jumper to install it into my unRAID server.  At first I got tons of errors when booting unRAID with the drives jumpered.  Then I tried zeroing the drives without and with the jumper on my Windows XP machine using the WD Windows Diagnostic Tool.  Zeroing the drive with the jumper allowed unRAID to boot without errors, but then errors would occur during the preclear process.  Funny thing though, preclear passed the drives, but the syslog was completely filled with read errors (over 2 gig of errors).  Maybe Joe should look into this just in case someone does not monitor the syslog when they preclear a disk.

After spending a week trying to get the drives to work in unRAID with the jumper, I cut my loss and RMA the drives.  I guess WD really means it when they say the drive cannot be jumpered once it has been formatted without the jumper.

I guess the guy selling jumpers on eBay might raise his price if he was reading this forum

Okay, that gives me something to think about.

I'm going to try the following.

1) Zero the drive without the jumper set using a DBAN boot disk

2) Install the jumper, then add it to unRAID, run pre-clear etc and see what happens

3) Give my brother-in-law a call, he works for Seagate, in the firmware development department, specifically storage array optimization for major storage companies, maybe he can shed some light on this

4) If the above yields no useful solutions, will RMA them, but one would think WD have some way of reseting these disk

5) If all else fails I will go and sit in the corner of my room, rocking and sucking my thumb....

I just want to report back on my failed attempt to properly set up the two WD15EARS drives I have.

I removed one disk on the devices page and powered down. I then put the jumper on and booted up. The server would not start (after 20min waiting). I disconnected the drive from the system and the server booted up.

Since my server is headless and I do not have a monitor, I used my LG HDTV to see what was going on. If the jumper is on and the disk is connected the screen showed errors like posted by the OP for about 10min and then got to a login prompt.

I booted the system from a USB drive with Ultimate Boot Disk on it to zero out the drive while it was still in the system (I'm lazy and don't have backplanes). All 4 of the wiping utilities would not even start. I just got some garbled text at the point where it looks like its iterating the drives. I assume there was a problem because the drives are connected to SATA cards. Maybe ACHI mode in the BIOS, I don't know.

I took the drive out and put it in a USB Dock connected to a Windows XP machine. I started the WD align tool and it said the drive was aligned correctly. Nothing to do there. Again I ran the Ultimate Boot Disk. The estimated time was around 27hrs and it was still at 0.0% after 30min (that can't be right). I stopped the process.

I would take all week to zero the drive, pre-clear with the jumper on, and then let the server rebuild the data. I still had to rebuild the data. I don't get any parity sync errors and the server runs flawlessly for months since I added those drives. It was late, I gave up.

Other thoughts:

1. Maybe I could have pre-cleared the drive (no zeroing first) when the server finally came up while connected to my TV.

2. Maybe the USB Dock was limiting in some way and I need to hook the drive to a SATA port on a motherboard to improve speed.

Well it turns out my brother-in-laws area of expertise is so specialized he couldn't offer a definitive answer.

I could be wrong here, but zeroing drive will not help if the drive has been low level formatted in some way, all you are doing is writing zero's to the underlying format, as I said, I not sure but does make some sense.

I going to try this - remember, my WD's were formatted and partitioned in XP without the jumper first.

1) without jumper - delete the partitions within XP

2) without jumper - zero them for the heck of it using UBCD

3) with jumper - attach them to unRAID and run pre-clear with fingers and toes crossed

Will be a few days before I get around to this but I'll let  you know the outcome.

I'll have another chat with my brother-in-law, he did say he will look into it, and to be fair, I'm not sure I explained it all that well.