Good morning,

 

I am a newbie on Unraid but love the product so far.  I am in the process of building my array and am adding drives as I transfer data from my windows based file server.  As a result, I am adding a drive at a time as I need to move data to Unraid first.  I searched the forum and it was suggested that I reboot the Unraid server (which has been done several times)

 

I am attempting to add a 250G Hitachi Drive to the array.  The clear operation is taking forever to complete.  The other drives in the array (2 seagates and 1 maxtor) completed in about an hour or so.  At this rate, the operation will take a day or two complete.  That doesn't make sense to me.  I think that the drive is good as it has worked fine on the windows server for about a year.

 

I have tried numerous reboots. 

 

Attached is the tail of the syslog and the settings for the drive.

 

Thanks for any help.

 

Kevin

 

 

From what I can see from your listing of the settings, your drive is NOT using UDMA mode.  That would result in much longer access times using PIO mode.  As you discovered, it will take much longer for the clear operation to complete if not using DMA.

 

If you were using DMA, the line would read as it does on my unRaid server (note the current value is "1" on my server and a "0" on yours):

using_dma              1              0              1              rw

 

Suspect the cabling to the drive first. First try re-seating it, then try a different IDE cable.  Make sure it is a newer style flat 80 pin cable and not one of the older 40 pin cables.  Round cables may look better to you, but often do not work well under Linux.

 

If the cable has several connectors for the drives and only one drive is plugged in, make sure the empty connector is the one in the middle and that the drive is plugged into the end connector.

 

Joe L.

 

Connect hard drive directly to pci card / mobo and make sure you are not using round cables... I had the exact same problems with round cables and I also ran into issues with mobile rack units.

Thanks for the help!!  I'll swap the cables when I get home from work and report the results.

3 cables (flat, 80 pin) later I still have the same results.  Any other ideas?

I also went into the mobo and set the drive to ultra dma5 just for kicks.  cat for the drive settings shows dma value =1.  Performance is absolutely terrible.

 

The part that is really odd is that the two drives on the secondary IDE controller were cleared without incident and they are larger drives (500G).

 

Any ideas on what to try next?? 

 

Hi KD,

 

I'd like to see the entire syslog immediately after a re-boot.

Also, what mobo are you using?

I have attached a syslog right after reboot.

 

Mobo is ASUS P5PE-VM, 2x512 Kingston Ram.  I am using the onboard ide controllers.  The two drives on the secondary IDE controller are functioning normally.  The primary controller is the one having problems (so far).

 

Thanks,

Kd

 

The troubling entry in the syslog is this:

 

Jan  7 11:48:37 Tower kernel: md: import hda HDS722525VLAT80 VNR98FC6CRGXAM offset: 63 size: 244198552
Jan  7 11:48:37 Tower kernel: md4: new disk
Jan  7 11:48:57 Tower kernel: hda: dma_timer_expiry: dma status == 0x21
Jan  7 11:49:07 Tower kernel: hda: error waiting for DMA
Jan  7 11:49:07 Tower kernel: hda: dma timeout retry: status=0x58 { DriveReady SeekComplete DataRequest }

 

If possible swap the m/b primary and secondary devices, that is, just unplug the cables from the m/b and swap them.  This will put the known working drives on primary and your suspect drive on the secondary channel.  You will need to make adjustment on the Devices page so that the disk numbers are the same.  Then reboot and see if this problem stays with the hard drive or the controller.

 

If you don't want to risk anything happening to your known good drives, another approach is this:

First, save a copy of the file 'config/super.dat' on the flash.  Next, remove power connector from all the known good drives and hook up your suspect drive by itself.  Try first the secondary channel (that you currently have it on), and then try the primary channel.  Go to the Devices page and assign the one drive as 'disk1' (no parity assigned).  When you start the system it will create single-disk array with no parity.  You should be able to format the disk and try writing stuff to it.  Check the syslog for errors.  Later when you want to restore your original config, you can hook up all the disks the way you had them, and then restore the copy of the 'config/super.dat' file.  Upon system restart, everything should be the way it was.

 

The advantage of this approach is that if the system works with just the one drive, then perhaps it's a power issue.

First - thanks for all the help.  One of the reasons that I went with Unraid was the responsive tech support and the active participation in the forums.  Very helpful.

 

In end there were two problems.  The first was the ide cable (thanks Joe L) and the second was the drive (Hitachi).  Changing both resulted in the configuration working as expected.  The drive just was not happy so I put it into a USB enclosure and reformatted it for use with one of the Windows PC's...

 

Thanks again for the help.