Hi.  I'm a total noob w/unRAID (and linux too for that matter) so I'm sure I probably just messed up something simple from the Configuration Tutorial but I couldn't find an answer searching the forums so thought I should post it here.  After running the recommended 3 preclear cycles on my new drives (3TB WD Reds) I assigned the 2 data drives, clicked to enable user shares, and then started the array.  (Was going to wait to assign the parity drive until after I transferred my data over.)  Anyway, so far so good.  Then I hit the button to format the drives and waited... and waited.  The tutorial says it should only take a few minutes to format precleared drives but almost an hour in disk 1 was still showing as "Formatting" and disk 2 as "Unformatted".  So then checking the syslog I see "Server emhttp: disk 1 mount error: 32" and the same line for disk 2 also.  Full syslog is attached if it helps.

 

So what did I do wrong and what next?  Thanks.

 

 

syslog_4-9-13.txt

Looks like the GPT partitioning took about a minute or two, but I see NO evidence of you formatting them with a reiserfs file-system.    Did you press the "Format" button?  (after checking the box under it to enable it?)  Never-mind, I now see it...  You started the formatting, but did not let it finish.

Apr  9 06:20:24 Server emhttp: writing GPT on disk (sda), with partition 1 offset 64, erased: 0

Apr  9 06:20:24 Server emhttp: shcmd (206): sgdisk -Z /dev/sda &> /dev/null

Apr  9 06:20:24 Server kernel: mdcmd (9): spinup 1

Apr  9 06:20:24 Server kernel: mdcmd (10): spinup 2

Apr  9 06:20:25 Server emhttp: shcmd (207): sgdisk -o -a 64 -n 1:64:0 /dev/sda |& logger

Apr  9 06:20:25 Server kernel:  sda: sda1

Apr  9 06:20:26 Server logger: Creating new GPT entries.

Apr  9 06:20:26 Server logger: The operation has completed successfully.

Apr  9 06:20:26 Server emhttp: shcmd (208): udevadm settle

Apr  9 06:20:26 Server kernel:  sda: sda1

Apr  9 06:20:26 Server emhttp: writing GPT on disk (sdb), with partition 1 offset 64, erased: 0

Apr  9 06:20:26 Server emhttp: shcmd (209): sgdisk -Z /dev/sdb &> /dev/null

Apr  9 06:20:27 Server emhttp: shcmd (210): sgdisk -o -a 64 -n 1:64:0 /dev/sdb |& logger

Apr  9 06:20:27 Server kernel:  sdb: sdb1

Apr  9 06:20:28 Server logger: Creating new GPT entries.

Apr  9 06:20:28 Server logger: The operation has completed successfully.

 

I takes much longer these days for the formatting that when disks were 500Meg.  It takes FAR longer when you are calculating parity at the same time.  Looks like you were too impatient.  To speed it up, cancel the parity calc, re-start it once the disks are formatted.

 

You'll probably need to re-issue the mkreiserfs commands

mkreiserfs -q /dev/md2

and

mkreiserfs -q /dev/md1

 

Joe L.

Looks like the GPT partitioning took about a minute or two, but I see NO evidence of you formatting them with a reiserfs file-system.    Did you press the "Format" button?  (after checking the box under it to enable it?)

Apr  9 06:20:24 Server emhttp: writing GPT on disk (sda), with partition 1 offset 64, erased: 0

Apr  9 06:20:24 Server emhttp: shcmd (206): sgdisk -Z /dev/sda &> /dev/null

Apr  9 06:20:24 Server kernel: mdcmd (9): spinup 1

Apr  9 06:20:24 Server kernel: mdcmd (10): spinup 2

Apr  9 06:20:25 Server emhttp: shcmd (207): sgdisk -o -a 64 -n 1:64:0 /dev/sda |& logger

Apr  9 06:20:25 Server kernel:  sda: sda1

Apr  9 06:20:26 Server logger: Creating new GPT entries.

Apr  9 06:20:26 Server logger: The operation has completed successfully.

Apr  9 06:20:26 Server emhttp: shcmd (208): udevadm settle

Apr  9 06:20:26 Server kernel:  sda: sda1

Apr  9 06:20:26 Server emhttp: writing GPT on disk (sdb), with partition 1 offset 64, erased: 0

Apr  9 06:20:26 Server emhttp: shcmd (209): sgdisk -Z /dev/sdb &> /dev/null

Apr  9 06:20:27 Server emhttp: shcmd (210): sgdisk -o -a 64 -n 1:64:0 /dev/sdb |& logger

Apr  9 06:20:27 Server kernel:  sdb: sdb1

Apr  9 06:20:28 Server logger: Creating new GPT entries.

Apr  9 06:20:28 Server logger: The operation has completed successfully.

 

I takes much longer these days for the formatting that when disks were 500Meg. Looks like you were too impatient.

 

Joe L.

 

Thanks Joe.  I would have left it running if I didn't see the "mount error: 32"'s in the syslog.  Sounds like I should just reboot and try again though so I'll do that when I get back home and just let it run this time.  (I'm sure I did click the Format button but I'll double check that I'm following those steps correctly too.)

 

Just to make sure I don't wait too long if the disks really aren't formatting though, can you give me any estimate on how long formatting the 2 data drives should take?  My server is a home build (my first) with an ASUS C-60M1-I board/cpu processor (AMD Fusion C-60 dual core 1.0 GHz APU) and each preclear cycle for the 3x3TB WD Reds took 46+ hours total, if that info provides any guide.

Looks like the GPT partitioning took about a minute or two, but I see NO evidence of you formatting them with a reiserfs file-system.    Did you press the "Format" button?  (after checking the box under it to enable it?)  Never-mind, I now see it...  You started the formatting, but did not let it finish.

Apr  9 06:20:24 Server emhttp: writing GPT on disk (sda), with partition 1 offset 64, erased: 0

Apr  9 06:20:24 Server emhttp: shcmd (206): sgdisk -Z /dev/sda &> /dev/null

Apr  9 06:20:24 Server kernel: mdcmd (9): spinup 1

Apr  9 06:20:24 Server kernel: mdcmd (10): spinup 2

Apr  9 06:20:25 Server emhttp: shcmd (207): sgdisk -o -a 64 -n 1:64:0 /dev/sda |& logger

Apr  9 06:20:25 Server kernel:  sda: sda1

Apr  9 06:20:26 Server logger: Creating new GPT entries.

Apr  9 06:20:26 Server logger: The operation has completed successfully.

Apr  9 06:20:26 Server emhttp: shcmd (208): udevadm settle

Apr  9 06:20:26 Server kernel:  sda: sda1

Apr  9 06:20:26 Server emhttp: writing GPT on disk (sdb), with partition 1 offset 64, erased: 0

Apr  9 06:20:26 Server emhttp: shcmd (209): sgdisk -Z /dev/sdb &> /dev/null

Apr  9 06:20:27 Server emhttp: shcmd (210): sgdisk -o -a 64 -n 1:64:0 /dev/sdb |& logger

Apr  9 06:20:27 Server kernel:  sdb: sdb1

Apr  9 06:20:28 Server logger: Creating new GPT entries.

Apr  9 06:20:28 Server logger: The operation has completed successfully.

 

I takes much longer these days for the formatting that when disks were 500Meg.  It takes FAR longer when you are calculating parity at the same time.  Looks like you were too impatient.  To speed it up, cancel the parity calc, re-start it once the disks are formatted.

 

You'll probably need to re-issue the mkreiserfs commands

mkreiserfs -q /dev/md2

and

mkreiserfs -q /dev/md1

 

Joe L.

 

Hi again.  Just to follow up since I see you edited your earlier reply, I hadn't assigned the parity disk yet so I wasn't calculating parity at the same time (or if I was I didn't mean to).  This was just formatting the 2 data disks. 

 

Also I didn't enter any mkreiserfs commands before (unless those are issued by some action I took through the server management utility?)  So maybe I missed a step?

 

If the fix is just to reboot, follow the steps in the configuration tutorial and be more patient w/the formatting this time then no need to give me more detail than that since it will probably just confuse me.    If I am missing a step though please let me know.  Thanks again.

Also I didn't enter any mkreiserfs commands before (unless those are issued by some action I took through the server management utility?)  So maybe I missed a step?

Those are the commands issued as the result of pressing the "Format" button.

If the fix is just to reboot, follow the steps in the configuration tutorial and be more patient w/the formatting this time then no need to give me more detail than that since it will probably just confuse me.    If I am missing a step though please let me know.  Thanks again.

You'll soon see.  If the disks do not mount, it is easiest to issue the two commands I listed than to get unRAID to start from the beginning.  It probably thinks they are corrupted disks right now.

 

Joe L.

Also I didn't enter any mkreiserfs commands before (unless those are issued by some action I took through the server management utility?)  So maybe I missed a step?

Those are the commands issued as the result of pressing the "Format" button.

If the fix is just to reboot, follow the steps in the configuration tutorial and be more patient w/the formatting this time then no need to give me more detail than that since it will probably just confuse me.    If I am missing a step though please let me know.  Thanks again.

You'll soon see.  If the disks do not mount, it is easiest to issue the two commands I listed than to get unRAID to start from the beginning.  It probably thinks they are corrupted disks right now.

 

Joe L.

 

Okay and I'll know if the disks don't mount if I get the same "mount error: 32"'s in the syslogs again?  Apologies if this is a dumb question but I'm still not following why I got that error in the first place or what it meant.  If it wasn't actually preventing the disks from formatting and I was just impatient then I'm fine just ignoring them next time.  But if that's what is telling me they're not mounting (sounds like it to me) then I'll try the commands you listed and start over.  It's just a little confusing for someone who's new to all this because there's nothing in the config tutorial about mounting disks -- just assigning, starting the array and clicking format, which is what I did. 

 

And if I do have to use the commands, is unRAID going to recognize the drives as precleared or would I need to preclear them again first (or just let them take forever to format)?  I'm not in any rush so I don't mind preclearing again.  I just want to make sure I'm doing this right.

Okay so when I booted last night I could see the 2 data drives still assigned but nothing that said either "formatted" or "unformatted" next to them after I started the array and no longer an option through the server management utility to format them either (the button was gone).  So I was going to do it through the mkreiserfs commands but then I thought I'd just check to see if I'd get more options through the server management utility if I went ahead and assigned the parity drive now too (since I was kind of on the fence about that before anyway).  Sure enough, as soon as I assign sdc as the parity drive I see a box come up to format and parity sync so I select that and let it run 5-6 hours to do the parity sync which completed successfully.  What I was hoping to see after though was some status on whether all the drives were no formatted correctly.  If what I did was a parity sync and format of sdc only and somehow sda and sdc still weren't formatted right (is that even possible?) then I'd expect to see something that tells me that.  But still none of the drives say anything about whether they're formatted or not.  So now I'm running a manual parity check to see if it might come up there.

 

Joe, would you mind taking a look at the new syslog attached and let me know if this looks right to you or if I still need to go back through the command line to format and then re-do the parity sync?  I appreciate your help.

syslog-2013-04-10.txt

Okay so when I booted last night I could see the 2 data drives still assigned but nothing that said either "formatted" or "unformatted" next to them after I started the array and no longer an option through the server management utility to format them either (the button was gone).  So I was going to do it through the mkreiserfs commands but then I thought I'd just check to see if I'd get more options through the server management utility if I went ahead and assigned the parity drive now too (since I was kind of on the fence about that before anyway).  Sure enough, as soon as I assign sdc as the parity drive I see a box come up to format and parity sync so I select that and let it run 5-6 hours to do the parity sync which completed successfully.  What I was hoping to see after though was some status on whether all the drives were no formatted correctly.  If what I did was a parity sync and format of sdc only and somehow sda and sdc still weren't formatted right (is that even possible?) then I'd expect to see something that tells me that.  But still none of the drives say anything about whether they're formatted or not.  So now I'm running a manual parity check to see if it might come up there.

 

Joe, would you mind taking a look at the new syslog attached and let me know if this looks right to you or if I still need to go back through the command line to format and then re-do the parity sync?  I appreciate your help.

everything looks normal now.

 

You can type

mount

on the command line and you'll see both your data disks are mounted.

 

You do not need to do anything now as far as I can see.  Parity is already in place, and the file-systems mounted.

 

Joe L.

 

Okay so when I booted last night I could see the 2 data drives still assigned but nothing that said either "formatted" or "unformatted" next to them after I started the array and no longer an option through the server management utility to format them either (the button was gone).  So I was going to do it through the mkreiserfs commands but then I thought I'd just check to see if I'd get more options through the server management utility if I went ahead and assigned the parity drive now too (since I was kind of on the fence about that before anyway).  Sure enough, as soon as I assign sdc as the parity drive I see a box come up to format and parity sync so I select that and let it run 5-6 hours to do the parity sync which completed successfully.  What I was hoping to see after though was some status on whether all the drives were no formatted correctly.  If what I did was a parity sync and format of sdc only and somehow sda and sdc still weren't formatted right (is that even possible?) then I'd expect to see something that tells me that.  But still none of the drives say anything about whether they're formatted or not.  So now I'm running a manual parity check to see if it might come up there.

 

Joe, would you mind taking a look at the new syslog attached and let me know if this looks right to you or if I still need to go back through the command line to format and then re-do the parity sync?  I appreciate your help.

everything looks normal now.

 

You can type

mount

on the command line and you'll see both your data disks are mounted.

 

You do not need to do anything now as far as I can see.  Parity is already in place, and the file-systems mounted.

 

Joe L.

 

Excellent!  Thanks Joe.  One more thing though.  Now that parity check has been running for an hour or so I'm seeing CPU stall errors galore in my syslog (see new log attached).  Looks bad to me so I was thinking I should stop the check but then I found this support thread with Tom saying that these are really just "harmless" warning messages. -- http://lime-technology.com/forum/index.php?topic=25425.msg221266#msg221266

 

Can you confirm that?  I don't want to be impatient and stop a process again if the parity check isn't hurting anything and I'm going to assume that it isn't and whatever these errors are I can just deal with them later (perhaps in a separate thread if you want me to mark this one as solved and move on).  Thanks again.

syslog-2013-04-10_2.txt