Jump to content

[SOLVED] Is there a way to re-install unRaid to get rid of this nasty problem?


Recommended Posts

Hey Guys,

 

I'm pretty much desperate, because there seems to be no solution for my problem:

 

http://lime-technology.com/forum/index.php?topic=15617.0

http://lime-technology.com/forum/index.php?topic=15638.0

 

No one knows what's the problem here, and I'm a complete Linux-illiterate..

 

Is there a way to re-install unRaid from the ground up, using my allready filled data disks 1 thru 8, hoping to get rid of this nasty problem??

 

Thanx!!

Link to comment

It sounds like you have a disk with issues. Look in the logs for a disk that is throwing errors and then look up reiserfsck in the Wiki and check your disks. I think the log might show md4 having the errors which corresponds to disk4 for example. If you can't figure out which disk then run reiserfsck on every data disk.

 

Peter

Link to comment

Ok,

 

That’s done.

I ran reiserfsck - -check on all my drives (except the Parity Disk).

All disks returned no defects..  :-\

 

I remembered when doing a Parity Check when the problem occured for the first time, the only faults I got were on the Parity Disk. But in the WiKi I read that I can not perform a reiserfsck on the Parity Disk, because that can damage the disk.

 

What should I do now, with no problems found on all drives?

Is there a way to just let the Parity Disk build up from scratch?

(Because I know there arn’t any problems on all other disks)

 

Maybe use the “Trust My Array” procedure??

Or am I thinking in the wrong direction?

 

Thanx!

Link to comment

Look at ID #5. That disk has 35 reallocated sectors. You could consider it OK if it can complete at least 2 or 3 parity checks without the sectors pending reallocation or the sectors reallocated increasing (ID#5 and #197). Personally, I would try to RMA it because you'll have big problems if another disk fails an it throws more bad sectors at the same time.

 

Peter

 

Link to comment

Ah allright,

 

The Parity Check also returned 68 errors on the Parity Drive.

I guess I should replace the disk with the errors.

 

Just to be sure: Is it my Parity Drive which is bad? Or is it an other drive? How can I see to which drive ID #5 belongs to?

 

I'll swap the bad one for a new one. Last week I ordered a 3TB WD anyway, it supposed to be in tomorrow.

 

I attached my new Syslog.txt and Smart report.

 

Thanx for all your help Peter!!

syslog.txt

smart.txt

Link to comment

ata5.0 is referring to a controller. I believe it ties to the parity following these lines which tie host 5 to disk0, which is the parity;

 

Oct  9 09:09:09 Tower emhttp: pci-0000:00:1f.2-scsi-0:0:0:0 host5 (sdf) WDC_WD2001FASS-00W2B0_WD-WMAY00568580

 

Oct  9 09:09:09 Tower kernel: md: import disk0: [8,80] (sdf) WDC WD2001FASS-0 WD-WMAY00568580 offset: 63 size: 1953514552

 

Peter

 

 

Link to comment

Allright Guys,

 

Completely lost again:

 

Allthough I built my 2 new 3TB disks in my rig, I could't get myself to really use one as Parity.

Especially because you guys warned me about the risks.

 

So I did this: I bought a new 2TB WD EARS disk to replace the broken Parity Drive (RMA, had to wait too long for the new one to arrive).

I let it build Parity, and everything seems just about fine.

(except for this problem I have now when powering up, the server doesn't boot by itself from the USB stick. I'll go and try to fix that myself before asking questions.)

 

But my problems start all over again when I want to use a User Share. ("Media")

I can make one, but it never fills up with the data which it supposed to show.

 

When stopping the array, all disks remain "unmounting", but never really do unmount.

Now I am on the same point as in the beginning of this problem.

 

Any ideas anyone? I can't use my server as I want to when I can't export all my data as if it is one big disk..

(Running on 4.7 now)

Syslog.txt

Link to comment

Allright Guys,

 

Completely lost again:

 

Allthough I built my 2 new 3TB disks in my rig, I could't get myself to really use one as Parity.

Especially because you guys warned me about the risks.

 

So I did this: I bought a new 2TB WD EARS disk to replace the broken Parity Drive (RMA, had to wait too long for the new one to arrive).

I let it build Parity, and everything seems just about fine.

(except for this problem I have now when powering up, the server doesn't boot by itself from the USB stick. I'll go and try to fix that myself before asking questions.)

 

But my problems start all over again when I want to use a User Share. ("Media")

I can make one, but it never fills up with the data which it supposed to show.

 

When stopping the array, all disks remain "unmounting", but never really do unmount.

Now I am on the same point as in the beginning of this problem.

 

Any ideas anyone? I can't use my server as I want to when I can't export all my data as if it is one big disk..

(Running on 4.7 now)

If the disks are "unmounting, then they are "busy" with files open, or processes using them as their current directory.

 

You MUST stop all add-on processes, and terminate any programs that are using them before shut down can occur.  

 

Walk first.   then run.

 

Stop running any add-ons.   Get you server so it can start and stop reliably, then add one at a time to see which is causing your issue.

 

Joe L.

Link to comment

It could be the Realtek LAN chipset on your motherboard. I assume this motherboard did work for a while with no issues and you didn't upgrade the unRAID version since that time???

 

I would also do this. Check each disk for shares (a share is any directory at the top level of each disk) which are the same name except the names have capitalization differences. Also, check on the flash drive for extra share config files - I believe they would be in a directory something like config\shares and they will have the same name as the share was called but given an extension. I can't check this until I get home but I believe the share config files are 2 directories deep on the flash. For both, delete any duplicates that don't belong. Heck, you can delete all the share config files if you want but you should delete the config files when the array is stopped.

 

Peter

 

Link to comment

My system worked fine from the start.

I never upgraded unRAID since then, thats right.

 

I did upgrade after my problem occured, but that was only last week.

 

So its not because of the upgrade that it's not working anymore.

 

I'll go and try your flash drive - share tips and get back when I'm done!

Link to comment

Ok,

 

Tried that.

First I stopped the array, then I removed all the files in the folder flash/config/shares.

Then started the array again.

 

I tried to make a new User Share "Media".

 

No result again, the share "Media" (which is immediatly visible in Windows Explorer) doesn't contain anything.. And again, after stopping the array, all disks remain "unmounting"..

 

Now there is a new file in the map /config/shares on the flash, named Media.cfg.

 

All folder names on all disks are exactly the same. Media (not media) on all disks as top level older, in there folders as Concerts, Disney, Documentaries, Movies, Pictures and TVseries.

 

Thanx anyway..

 

More ideas?

Link to comment

unRAID should just pick the Media directories off the disks and create the user share when the array is started.

 

You could try re-installing. Take a screen capture of the main unRAID interface page so you can re-assign the disks correctly. Copy the key file to your PC - the file Limetech sent you for the license, I believe it'll be called pro.key or something like that. Then, just re-format the USB drive and re-install unRAID as per the instructions.

 

Then, put the key file back on the flash, boot up unRAID and re-assign the disks the same as before. unRAID will want you to build the parity again which you could cancel until you get the shares setup. You will likely have to turn-on user shares and then stop and start the array at which point the Media share should just appear by itself. Once it's working, start the parity build again and let it complete.

 

Peter

 

Link to comment

Odd.

 

Did you try stopping and restarting the array a few times?

Did you try re-creating the user share one last time?

 

I see these lines in your syslog;

 

Oct 17 23:38:00 Tower emhttp: shcmd (123): mkdir /mnt/user

Oct 17 23:38:00 Tower emhttp: shcmd (124): /usr/local/sbin/shfs /mnt/user -cache 1 -floor   -o noatime,big_writes,allow_other,default_permissions

Oct 17 23:38:00 Tower emhttp: _shcmd: shcmd (124): exit status: 1

 

This is where it is trying to create the mount point for the user share. I believe the exit status of 1 means it failed to create the /mnt/user point but I have no idea why that would fail.

 

I then see these lines;

 

Oct 17 23:38:00 Tower emhttp: shcmd (126): /usr/local/sbin/shfs /mnt/user0 -cache 0  -o noatime,big_writes,allow_other,default_permissions

Oct 17 23:38:01 Tower emhttp: shcmd (127): killall -HUP smbd

Oct 17 23:38:01 Tower emhttp: shcmd (128): /etc/rc.d/rc.nfsd restart | logger

 

This is where it creates the user0 mount point and then I believe the last 2 lines are to re-start samba with these newly created user share mount points being actually network shared.

 

Logon to the server and do the following;

 

cd /mnt

ls

 

and you should see directories user and user0, which hold the user shares. user is the whole share and user0 is the share without the contents of the cache disk. I'm betting that user is missing.

 

I would try to unassign the cache disk and re-start the array. Just a guess, but it appears that user, which includes the cache, isn't being created while user0, which doesn't include the cache, is working so I'm thinking it may be a cache drive issue of some sort.

 

If user is missing then you may want to email tomm at limetech.com and ask for help directly at this point.

 

Peter

 

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...