Share is gone - only Data from Cache drive


raVensen2k

Recommended Posts

Hello Community, 

i need some help here. Following situation:

I have a unraid pc working as a nas. The Nas is running with 3x 4TB WD RED (2x Data, 1 Parity) and 1 500gb SSD as Cache. I used about 3 TB of the 8TB available.

The NAS went fine for a while. Then i moved to another city and lost my usb-stick with unraid while moving. 

Yesterday i made a new usb-stick with an usb-backup from february (yeah, i know, its old). i got myself a new registration key for the new usb-stick and everythink went fine. i got access to my data and i was happy. The network drive showed the full 8 TB partition. The backup was with Unraid 6.8.2 and i upgraded imediatly to 6.8.3, rebooted and went to bed.

Today i wanted to copy some files to my NAS and noticed in my Windows Explorer from my computer that the network drive from the NAS is only 465gb big. Most of the data was missing.

Of course i didnt went to the forum instantly and read anything about it, instead i rebootet the nas. After the reboot my share (NAS) was missing completly.

After that i re-flashed the new usb-stick with the backup from february, reinstalled the new key which i got yesterday and bootet again. But now i only see the data which is on my cache file.

 

Since i got access to my files yesterday, i did nothing with the nas as upgrading it from 6.8.2 to 6.8.3. i have no clue whats going on and if i lost all my data (that would hurt a lot).

Attached is my syslog-zip, which i created after reading the post: need help? read me first!

 

Do you guys have any idea whats going on and if i can get my data back? what do i need to do?

 

In the Main window you can see that my HDDs are nearly empty now. As i wrote i was using about 3TB.

image.thumb.png.74e42a57b0f2c7596767382cb5d89bae.png

Here some more screenshots i think could be usefull

image.thumb.png.df5fb9c6b698f3a685cd968bfb71fe62.png

image.thumb.png.ecea5d28aab8145254c57cbb87355b89.png

 

 

image.png

image.thumb.png.f34527a1b0242ebb1b4caed7fc1cefae.png

nas-syslog-20200918-1735.zip

nas-diagnostics-20200918-2012.zip

Edited by raVensen2k
Link to comment
2 hours ago, raVensen2k said:

if i lost all my data (that would hurt a lot)

Do you have backups of anything important and irreplaceable? Parity is not a substitute for a good backup plan. Lots of more common ways to lose data besides failed disks, including user error.

 

2 hours ago, raVensen2k said:

my HDDs are nearly empty now.

Are you sure you didn't format them, perhaps because they were unmountable? Formatting would be the most likely cause.

 

Syslog resets on reboot so can't see what happened before.

 

Are these the only plugins you normally run?

community.applications.plg - 2019.12.21
rclone.plg - 2019.11.01
user.scripts.plg - 2019.08.17

Looks like CA and User Scripts are out-of-date. Not likely the problem but it does make me wonder if you might have had other out-of-date plugins that you removed.

 

2 hours ago, raVensen2k said:

The network drive showed the full 8 TB

It is still showing you have total array storage of 8TB, just not much used.

 

2 hours ago, raVensen2k said:

made a new usb-stick with an usb-backup from february

There are some details at this point that might provide some clues if you can elaborate.

 

Did you have to rebuild parity when you started with the new flash drive?

 

Are you absolutely sure the disks were assigned to the correct slots before you started the array?

Link to comment

Also, lots of these at the end of your syslog

Sep 18 10:33:06 NAS kernel: BTRFS info (device sdf1): relocating block group 514199781376 flags data|raid1
Sep 18 10:33:10 NAS kernel: BTRFS info (device sdf1): found 24 extents

Not sure what they are about but not normal.

 

Was there anything related to cache you had to do when starting with a new flash?

 

appdata, domains, system shares have files on the array instead of all on cache where they should be. In fact, for some reason you have appdata set to cache-no instead of cache-prefer. And I don't know what the nas share is used for or why you would want it all on cache.

 

These things may be completely unrelated to your data loss, just some things I noticed in diagnostics. Don't know if there are any clues among them or not.

Link to comment

Hello,

sorry for my late reply, i have been away for a few days. And now there is a lot to answer, so here we go.:

On 9/18/2020 at 10:37 PM, trurl said:

Do you have backups of anything important and irreplaceable? Parity is not a substitute for a good backup plan. Lots of more common ways to lose data besides failed disks, including user error.

My backup is pretty much not existent. i know thats fatal. the only thing which could safe my ass is the fact that i have my old raid controller with the old hdds laying around and i hope these still works. if that is the case i could get back about 90% of my data.

 

On 9/18/2020 at 10:37 PM, trurl said:

Are you sure you didn't format them, perhaps because they were unmountable? Formatting would be the most likely cause.

Yes, i am sure i didnt format them.

 

On 9/18/2020 at 10:37 PM, trurl said:

Looks like CA and User Scripts are out-of-date. Not likely the problem but it does make me wonder if you might have had other out-of-date plugins that you removed.

Some time ago i found an article to integrate my google drive into unraid. i tried that and tried with scripts to backup my data from unraid to the mounted google drive. but that didnt work because i have no clue about scripts.

 

On 9/18/2020 at 10:37 PM, trurl said:

It is still showing you have total array storage of 8TB, just not much used.

No, its not and that is strange. Now it only shows 232GB in Windows as max capacity which is the size of the cache drive. When i open the network storage i find all the files and links but nearly everything has 0kb size.

 

On 9/18/2020 at 10:37 PM, trurl said:

Did you have to rebuild parity when you started with the new flash drive?

After i bootet the first time with the new usb-stick i saw all my data with the 8TB network drive NAS. I was able to copy all my pictures to my local PC.

Unraid wanted to check parity, but i didnt saw it rebuilding.

On 9/18/2020 at 10:37 PM, trurl said:

Are you absolutely sure the disks were assigned to the correct slots before you started the array?

I didnt swap any cables since i build this nas. these cables are in the same position, thats for sure.

 

On 9/18/2020 at 10:45 PM, trurl said:

Was there anything related to cache you had to do when starting with a new flash?

 

appdata, domains, system shares have files on the array instead of all on cache where they should be. In fact, for some reason you have appdata set to cache-no instead of cache-prefer. And I don't know what the nas share is used for or why you would want it all on cache.

Im not very sure at the moment to be honest. I think i fired it up and it was ok, i saw that only one cache-drive was assigned to the raid. i stopped the array and added the second cache drive.  Now what i see is that my "NAS" share is only on the Cache drive and not on the big drives anymore. Thats not right. 

I have no idea what the settings were where appdata, domains, system and so on should have been. I did not touch these settings when i installed unraid. i only set up my array, users and smb share.

 

Now it seems that everything swaped. my cache drive is the "NAS" drive now where all my data should be and on my 8TB share are the appdata, domains .... 

I made a few screenshots: Nas share is only 232GB big, as the cache drive. On the 4TB HDD Disk 1 is only appdata, domains, isos, system, but not nas.

On disk 2 is only appdata.

 

I hope i answered all the questions. i think i fucked it up, right? ;)

image.thumb.png.390adeb6f17384485b7c2f45f60fe660.png

image.thumb.png.73c453834da011a9b11a9ca2977aac33.png

Link to comment
4 hours ago, raVensen2k said:

I didnt swap any cables since i build this nas. these cables are in the same position, thats for sure.

cables and ports are unrelated to my question since Unraid keeps track of disk assignments by drive serial number. I was asking about the disk assignments and whether or not you may have assigned them incorrectly. Accidentally assigning a data disk to the parity slot will overwrite that data with parity. Possibly all the missing data was on disk1, for example, and you accidentally assigned that disk as parity.

Link to comment
4 hours ago, raVensen2k said:

o, its not and that is strange. Now it only shows 232GB in Windows as max capacity which is the size of the cache drive. When i open the network storage i find all the files and links but nearly everything has 0kb size.

If you ever tried to copy between a disk share and a User share you can end up with all your files truncated to 0 size.    This is because a User Share is just another view of the files on the Disk Shares, and when you do such a copy you end up trying to write the file over-itself which ends up truncated to 0 size.  The reason is an architectural one in that because the paths are different the Linux level does not realise they are actually the same file so does not block the copy attempt.   This is one of the reasons that Disk Shares are not normally enabled by default.

Link to comment
1 hour ago, trurl said:

cables and ports are unrelated to my question since Unraid keeps track of disk assignments by drive serial number. I was asking about the disk assignments and whether or not you may have assigned them incorrectly. Accidentally assigning a data disk to the parity slot will overwrite that data with parity. Possibly all the missing data was on disk1, for example, and you accidentally assigned that disk as parity.

Hello, thank you for your reply. 

i didnt know that unraid works like this. but there was no need for any reajustment of the array. after i bootet from the new usb drive the first time, the assignment of the drives seemed fine. 1x Parity, 2x Data, 1x Cache was already set up and with this config i started the array.

 

 

Link to comment
1 hour ago, itimpi said:

If you ever tried to copy between a disk share and a User share you can end up with all your files truncated to 0 size.    This is because a User Share is just another view of the files on the Disk Shares, and when you do such a copy you end up trying to write the file over-itself which ends up truncated to 0 size.  The reason is an architectural one in that because the paths are different the Linux level does not realise they are actually the same file so does not block the copy attempt.   This is one of the reasons that Disk Shares are not normally enabled by default.

If i see this correctly i never had a disk share, only user shares.

Link to comment
9 minutes ago, raVensen2k said:

If i see this correctly i never had a disk share, only user shares.

A better way to say it is you shouldn't mix disks and user shares when copying / moving. If you are not sharing disks on the network then you don't have disk shares, but you still have disks, 

 

If you work directly with the disks on the server, such as with krusader / mc / command line, and you move or copy from disks to user shares or from user shares to disks, you can wind up with empty files. This is because linux doesn't know a disk file might be the exact same file as a user share file if the paths work out that way. So it tries to write what it needs to read, starting with creating the empty file to be written into.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.