Don't get me wrong, the system isn't frozen.

Simply, on the main page, all the disk of my (new-old) server are "formatting", but the reads and write counters are stopped.

 

A little story : I'm currently assembling a server with six hard drive 1TB, that where in another unraid server before.

So, I first preclear all those drives one time, then I assign all the drives in one single operation (5x data + 1x parity).

After reboot, all the drives (except parity) were blues, and I let unraid doing the formatting. I think it was creating the parity.

Now, and from a while, the drives are labelled as formatting (except parity). The balls are green, but blinking (so the drives are stopped, as spindown stated in the syslog).

There are no data at all in this system, and I think I've made a mistake, but what exactly I don't know. It may be not a good idea to let the system concurrently calculate the parity while all the drives are formatted?

Anyway, all the drives should be quickly formatted, the succesfull preclear was just done.

Now, I'm stuck, I can of course shutdown the system, initconfig, and so one, but is it the good solution? Something more appropriate?

 

n.b. The system is not the system in my signature, so a quick description : 6x 1TB seagate on MB, unraid 4.7.

 

syslog-2011-09-22_stick_formatting.txt

Don't get me wrong, the system isn't frozen.

Simply, on the main page, all the disk of my (new-old) server are "formatting", but the reads and write counters are stopped.

 

A little story : I'm currently assembling a server with six hard drive 1TB, that where in another unraid server before.

So, I first preclear all those drives one time, then I assign all the drives in one single operation (5x data + 1x parity).

After reboot, all the drives (except parity) were blues, and I let unraid doing the formatting. I think it was creating the parity.

Now, and from a while, the drives are labelled as formatting (except parity). The balls are green, but blinking (so the drives are stopped, as spindown stated in the syslog).

There are no data at all in this system, and I think I've made a mistake, but what exactly I don't know. It may be not a good idea to let the system concurrently calculate the parity while all the drives are formatted?

Anyway, all the drives should be quickly formatted, the succesfull preclear was just done.

Now, I'm stuck, I can of course shutdown the system, initconfig, and so one, but is it the good solution? Something more appropriate?

 

n.b. The system is not the system in my signature, so a quick description : 6x 1TB seagate on MB, unraid 4.7.

 

 

I know this might sound crazy, but is the array started? (it should be)

 

if yes, post a syslog.  There have been many who had the server deadlock when attempting to format a large number of drives at one time.

I know this might sound crazy, but is the array started? (it should be)

 

if yes, post a syslog.  There have been many who had the server deadlock when attempting to format a large number of drives at one time.

 

An illustration is sometime better than an explanation :

 

 

 

The disc balls are gray at the time of capture (but blinking green).

 

The syslog was already on the first post.

 

Edit : look at this, strange isn't?

 

looks a lot like it is sitting there deadlocked.

I can see all the format commands in the syslog.

 

Sep 22 14:30:43 Tour emhttp: shcmd (40): mkreiserfs -q /dev/md3 2>&1 | logger

Sep 22 14:30:43 Tour emhttp: shcmd (40): mkreiserfs -q /dev/md4 2>&1 | logger

Sep 22 14:30:43 Tour emhttp: shcmd (40): mkreiserfs -q /dev/md2 2>&1 | logger

Sep 22 14:30:43 Tour emhttp: shcmd (40): mkreiserfs -q /dev/md1 2>&1 | logger

Sep 22 14:30:43 Tour emhttp: shcmd (40): mkreiserfs -q /dev/md5 2>&1 | logger

 

You could try

ps -ef | grep mkreiserfs

to see if they are still running

You could try

ps -ef | grep mkreiserfs

to see if they are still running

 

 

Tour login: root

Password:

Linux 2.6.32.9-unRAID.

root@Tour:~# ps -ef | grep mkreiserfs

root      2004  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md3 2>&1 | logger

root      2005  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md4 2>&1 | logger

root      2006  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md2 2>&1 | logger

root      2007  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md1 2>&1 | logger

root      2008  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md5 2>&1 | logger

root      2009  2008  0 14:30 ?        00:00:21 mkreiserfs -q /dev/md5

root      2011  2005  0 14:30 ?        00:00:00 mkreiserfs -q /dev/md4

root      2013  2004  0 14:30 ?        00:00:00 mkreiserfs -q /dev/md3

root      2015  2007  0 14:30 ?        00:00:21 mkreiserfs -q /dev/md1

root      2016  2006  0 14:30 ?        00:00:20 mkreiserfs -q /dev/md2

root      6867  6855  0 22:40 pts/0    00:00:00 grep mkreiserfs

root@Tour:~#

 

 

And now? Powercycle and initconfig? 

You could try

ps -ef | grep mkreiserfs

to see if they are still running

 

 

Tour login: root

Password:

Linux 2.6.32.9-unRAID.

root@Tour:~# ps -ef | grep mkreiserfs

root      2004  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md3 2>&1 | logger

root      2005  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md4 2>&1 | logger

root      2006  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md2 2>&1 | logger

root      2007  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md1 2>&1 | logger

root      2008  1499  0 14:30 ?        00:00:00 sh -c mkreiserfs -q /dev/md5 2>&1 | logger

root      2009  2008  0 14:30 ?        00:00:21 mkreiserfs -q /dev/md5

root      2011  2005  0 14:30 ?        00:00:00 mkreiserfs -q /dev/md4

root      2013  2004  0 14:30 ?        00:00:00 mkreiserfs -q /dev/md3

root      2015  2007  0 14:30 ?        00:00:21 mkreiserfs -q /dev/md1

root      2016  2006  0 14:30 ?        00:00:20 mkreiserfs -q /dev/md2

root      6867  6855  0 22:40 pts/0    00:00:00 grep mkreiserfs

root@Tour:~#

 

 

And now? Powercycle and initconfig? 

looks like they are all running (and apparently deadlocked)

Thanks for your answers, Joe.

 

After a clean powercycle (power button) I've tried a second time, and after a while, same behaviour.

 

Restart, but this time I've stopped the parity check before format, and then everything works correctly. The server was up and ready after no time.

Then I made a successfull parity check.

Solution : don't let a parity check while formatting (in this case, all the drive).