The preclear of my second spare disk finished so I decieded to add one of the precleared disks top the array.  I stopped the array, used the  device management page to add the new disk (sdd) I started the array.

 

It then told me that the two data disks are unformatted. Also I can not see the two data disks in the network only the flash drive.

 

I started a parity check to see what it says

 

I noticed the following in the Syslog which does not look good.

 

Apr 19 16:02:11 Tower emhttp: shcmd (14): mkdir /mnt/disk1

Apr 19 16:02:11 Tower emhttp: shcmd (14): mkdir /mnt/disk2

Apr 19 16:02:11 Tower emhttp: shcmd (15): mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md1 /mnt/disk1 >/dev/null 2>&1

Apr 19 16:02:11 Tower emhttp: shcmd (16): mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md2 /mnt/disk2 >/dev/null 2>&1

Apr 19 16:02:11 Tower emhttp: _shcmd: shcmd (16): exit status: 32

Apr 19 16:02:11 Tower emhttp: disk1 mount error: 32

Apr 19 16:02:11 Tower emhttp: shcmd (17): rmdir /mnt/disk1

Apr 19 16:02:11 Tower emhttp: _shcmd: shcmd (17): exit status: 32

Apr 19 16:02:11 Tower emhttp: disk2 mount error: 32

Apr 19 16:02:11 Tower emhttp: shcmd (18): rmdir /mnt/disk2

Apr 19 16:02:11 Tower kernel: mdcmd (19): check

Apr 19 16:02:11 Tower kernel: md: recovery thread woken up ...

Apr 19 16:02:11 Tower kernel: md: recovery thread has nothing to resync

 

 

Any help greatly appreciated

The preclear of my second spare disk finished so I decieded to add one of the precleared disks top the array.  I stopped the array, used the  device management page to add the new disk (sdd) I started the array.

 

It then told me that the two data disks are unformatted. Also I can not see the two data disks in the network only the flash drive.

 

I started a parity check to see what it says

 

I noticed the following in the Syslog which does not look good.

 

Apr 19 16:02:11 Tower emhttp: shcmd (14): mkdir /mnt/disk1

Apr 19 16:02:11 Tower emhttp: shcmd (14): mkdir /mnt/disk2

Apr 19 16:02:11 Tower emhttp: shcmd (15): mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md1 /mnt/disk1 >/dev/null 2>&1

Apr 19 16:02:11 Tower emhttp: shcmd (16): mount -t reiserfs -o noacl,nouser_xattr,noatime,nodiratime /dev/md2 /mnt/disk2 >/dev/null 2>&1

Apr 19 16:02:11 Tower emhttp: _shcmd: shcmd (16): exit status: 32

Apr 19 16:02:11 Tower emhttp: disk1 mount error: 32

Apr 19 16:02:11 Tower emhttp: shcmd (17): rmdir /mnt/disk1

Apr 19 16:02:11 Tower emhttp: _shcmd: shcmd (17): exit status: 32

Apr 19 16:02:11 Tower emhttp: disk2 mount error: 32

Apr 19 16:02:11 Tower emhttp: shcmd (18): rmdir /mnt/disk2

Apr 19 16:02:11 Tower kernel: mdcmd (19): check

Apr 19 16:02:11 Tower kernel: md: recovery thread woken up ...

Apr 19 16:02:11 Tower kernel: md: recovery thread has nothing to resync

 

 

Any help greatly appreciated

As long as you did not press the "Format" button you are fine.  Just Stop the array and re-start it.

The drive should mount ok.

 

The parity check is at a lower level than the file systems.  It will not do anything to help mount the disks, you can cancel it.

 

This is the bug everybody is talking about in 4.5.3.  disks do not initially mount, and and disk that does not mount is incorrectly show as "Unformatted" 

 

If you press the Format button, unRAID will format the drives... DO NOT PRESS IT until only the new drive shows as unformatted.

Stop/Cancel the parity check.

Stop the array.

Start the array.

 

What happens then?

 

Write to lime-tech, describe the bug... point them to this thread.  Request it be fixed!!! Too many have already pressed the "Format" button in error losing their data.

 

Joe L.

Stop/Cancel the parity check.

Stop the array.

Start the array.

 

What happens then?

 

You should see the disks with file systems mount properly, and the new disk show as unformatted.

 

If you see just the new disk as unformatted, then it is OK to format it.  otherwise, do not press the "Format" button. Stop and re-start the array once more  (do not reboot, just stop the array, and then press Start).

DAMMM My impatience!!!!!

 

I did the stop start thing.  Disk1 came up fine.

 

Disk 2 which has no data on it first said mounting then after a while I hit refresh and it said unformatted..  I was under the impression that the preclear would write the signature so that when the disk was added to the array it would not need to format it. At least the preclear told me the disk was in good shape.

 

Well (sorry Joe) I pressed the format button as it indicated it will only format unformatted disks. Hopefully that will be okay and it will do its thing to the new disk (Disk2) and leave disk1 alone.

 

All hopefully is not lost, but "all" is not too much.  I will let y'all (pardon the pun, couldn't resist) know what happens.

 

I have got to learn to be more patient.  After all you guys replied in under 5 minutes!!!!  Anyone who bitches about support from this community should have their head examined!!

Okay, I did not press the format button until only the new disk said unformatted.  Thanks to whomever watches over me (probably my late wife who used to laugh at my impatience). 

 

I suspect I am okay.  I will send this to lime technology.  I assume I just need to send them an email as I saw no official bug reporting system, and point them tothis thread

I assume I just need to send them an email as I saw no official bug reporting system, and point them tothis thread

[email protected]

 

I am now out of panic mode.  It seemed the format finished in a few minutes and all looks totally normal.  Both disks are visible on the network.  I can see all the data that was on disk 1 and disk 2 reports empty as it should. Is that what should happen? i.e. the newly added disk should state unformatted and all th prior array disks show their normal unused size?

 

Thanks again to all for lowering my blood pressue.  

 

All is well again in Utopia (by the way I live in Utopia; it is the name of the street I live on.).  If any of this community is ever around Merritt Island FL drop in for a drink.  I get visitors from all over the world passing through because of contacts I have made through the WWW. Directions are on the web site at www.the-gordons.net

The pre-clear script clears a disk, and it partitions the disk, but it does not format it. (If it did, it would not be all zeros)

 

When you add a new disk, it will always attempt to mount it.  If the mount fails, it will show as unformatted and present the "Format" button.

 

On a new disk, the mount is expected to fail.  On an existing disk, the "mount" is NOT expected to fail, but on 4.5.3 it is failing at times for some users.  (I've never seen it on my older array, so it might be timing out too fast on users with newer hardware...) I don't honestly know the underlying cause, other than users reported stopping and re-starting the array let it properly mount the disk, so I suspect a timing issue.

 

several users did press the "Format" button when all their disks failed to mount.  It was not an easy situation to recover from.   This is about the most serious bug I've seen with unRAID in the 4+ years I worked with it.

 

Joe L.

I am now out of panic mode.

Thanks again to all for lowering my blood pressue.

A thank you email to Limetech is still a good idea.

 

I am now out of panic mode.

Thanks again to all for lowering my blood pressue.

A thank you email to Limetech is still a good idea.

Don't you mean a "Please fix this serious bug email to Limetech is a good idea." ??

Perhaps he meant 'Thank you for exercising my autonomic nervous system...'

 

I dunno, Purko and I are getting on eerily well today...

I fear an imminent earthquake... or tornado at least.

 

Wow, glad I didn't upgrade from 4.5.1. This makes me sort of doubt if unraid is really the right move... Is developing/maintaining/supporting this thing more than one man can handle ?

 

Do we have any possibility to post a big sticky in the top of the forum to warn users of this bug ?

I've read of a few cases were v4.5.3 won't mount some disks

/the array and by stopping and starting the array again fixes this issue. At one time I had v4.5.3 but rolled back to v4.5.1 because of the WOL failing to work and the S3 feature inconjunction with the S3 script wasn't spinning down my disks when it was suppose to either. It was an update that Tom had to release for the sake of supporting additional SATA RAID cards to replace other cards which were in short demand and needed this update to roll out for the servers which he builds and sells (I stand to be corrected but I believe this was the case).

This update seems to have brought this issue along with it. I wouldn't be put off by unraid just for this problem, if you're hardware is supported by v4.5.1, just roll back, problem solved. If we and the rest of Microsofts customers were put off by the amount if times Microsoft brought out a dodgy patch, Bill Gates would be a broke man and Microsoft would of been long gone by now.

 

Wow, glad I didn't upgrade from 4.5.1. This makes me sort of doubt if unraid is really the right move... Is developing/maintaining/supporting this thing more than one man can handle ?

 

Do we have any possibility to post a big sticky in the top of the forum to warn users of this bug ?

OK, refresh my memory - name one M$ bug that could cause you to loose all your data without the involvement of malware/hackers.

No? This is the big one, the mother of all bugs.

 

Well the data isn't lost unless you hit the 'Format' button, and if by accidently do this the data can still be restored by running a reiserfs file restore commands worse case scenario. Windows ME, Window Vista have had updates which cause instabilities and even when they we're released out of the box were plagued with problems which resorted to data loss for some not knowing what to do properly to get their data back.

Yes 4.5.3 has is a bug, but just roll back or stop and start your array if it occurs. Simple .

 

OK, refresh my memory - name one M$ bug that could cause you to loose all your data without the involvement of malware/hackers.

No? This is the big one, the mother of all bugs.

 

Well the data isn't lost unless you hit the 'Format' button, and if by accidently do this the data can still be restored by running a reiserfs file restore commands worse case scenario. Windows ME, Window Vista have had updates which cause instabilities and even when they we're released out of the box were plagued with problems which resorted to data loss for some not knowing what to do properly to get their data back.

Yes 4.5.3 has is a bug, but just roll back or stop and start your array if it occurs. Simple .

Yes, simple indeed.  Just tell that to the people who lost files in the past month.  And how did we end up comparing with windows?

 

well in the tone that i worded it, it wouldn't be good to happen, but it is something that can be fixed and i have to agree it is a serious fault too. . the windows analogy was just an example.

...but it is something that can be fixed...

Many of those who got their disks formatted were able to partly recover their files thanks to helpful forum members.  But I can't call that 'fixed'.

If you seriously claim that "it is something that can be fixed", you won't mind formatting your disks as a demonstration, will you?

 

If it is a problem for those who haven't had data loss occur to them, go back to v4.5.1, and don't keep v4.5.3.

If you or those who have a problem with this version and for those whose had data loss happen to them because of this problem, don't direct your anger about this problem to me, direct it to Lime Technology.

Yes I have to admit I should used a better choice of words on my earlier post and it wasn't the best way to put it.

Yes I would and others in this situation if data loss were to occur would be in panic mode and would be very angry if it was related to a bug like this and to no fault of my own, but it is not the end of the world and can be restored if it were to happen.     

 

...but it is something that can be fixed...

Many of those who got their disks formatted were able to partly recover their files thanks to helpful forum members.  But I can't call that 'fixed'.

If you seriously claim that "it is something that can be fixed", you won't mind formatting your disks as a demonstration, will you?

 

As a new user that has actually encountered the bug, it was not a disaster for me.  It was more or less after a few minutes "an interesting situation".  Initially it was panic mode, but after 10 minutes and a little forum help it went into the interesting class.

 

Lets all understand something.  Very few things based on general purpose computers are appliances.  Heck, even some "appliances" are not appliances, although I will admit that most of the non appliance appliances have a computer of some sort and software to define and control them.

 

If an unRaid user is totally unfamiliar with the word format and its various incarnations I would respectfully suggest they stay away from unRaid and the use of computers in embedded and sophisticated environments.  

 

There is no doubt in my mind that this is a serious bug.  In fact there should be, IMHO, an unRaid option for user level where a beginner gets querried for confirmation (hopefully with an explanation, perhaps written by community members). An advanced user might just get their commands carried out.

 

When I saw all my disks come up as unformatted I did not immediately just try and format them.  I stopped the array and restarted it and then just the newly precleared and added disk showed as unformatted.  I then went to our community and started this thread..  No data loss, a reasonable amount of caution, and some common sense.

Yes I would and others in this situation if data loss were to occur would be in panic mode and would be very angry if it was related

to a bug like this and to no fault of my own, but it is not the end of the world and can be restored if it were to happen.

Just why do you keep insisting that it can be restored? 

In all the related posts I see reports they restored most of the files, but nobody said "I restored all files, it wasn't the end of the world."

 

If the data you are storing is not the only copy then you should be able to restore it from the original copies. In that scenario all can be restored.  Might be very time consuming and a PITA but clearly doable.

 

I intend to keep all of the BR and DVD disks I rip.  If I sold or gave away the disks then I ethically feel I would be breaking the law.  As It stands now the laws in this matter appear to be in conflict in the USA. It is Fair Use vs Digital Rights Management.  I stand on the side of Fair Use.