I want to preface this with the fact that I am very early in my unraid interactions. I setup my system to take advantage of a windows rig I built in order to take over for a name brand NAS box and hopefully run some vms/ docker. I have had a stable system running basics of storage, some plex transcoding, and a VM or two with no stability issues or concerns.

 

I am in a position right now where:

- If I setup a new VM - the settings won't be remembered

- My docker containers are consistently reset or want to be setup whenever anything is opened

- Considerable amount of parity errors

- I lose connection to my admin view/unraid portal but then it is restored minutes later with no change in what unraid is working on - just lose connectivity??

 

Not sure what has changed and would really appreciate your help. I dont want to migrate additional data until I can figure out what is going on.

 

Unraid version: 6.9.2

CPU: Ryzen 5 2600

RAM: 8x4 G Skill RipJaw 3200GhZ

MotherBoard: Asus TUF gaming X570-plus Wi-Fi

Power Supply: EVGA 550 G3

Graphics Card: MSI RX570 8GB

 

Drives: 

Cache:

500GB Samsung 850 SSD

500GB Samsung 860 SSD

 

Parity:

12TB WD White

 

Others:

2TB Toshiba HDD --- starting to get indicators on

1TB Samsung 860 SSD

4TB WD Red

 

Thank you for your time and help. 

syslog.txt

Please post the entire diagnostics

 

5 hours ago, Squid said:

Please post the entire diagnostics

 

Thanks for the additional direction - adding now.

fractal-diagnostics-20211126-1223.zip

Your cache looks mostly empty. That and the fact that appdata and system shares have files on the array suggests you had cache problems at some point, which would explain losing dockers/VMs to some extent. Can you tell us anything about that? 

 

30 minutes ago, trurl said:

Your cache looks mostly empty. That and the fact that appdata and system shares have files on the array suggests you had cache problems at some point, which would explain losing dockers/VMs to some extent. Can you tell us anything about that? 

 

 

Oh good point - I had a 120GB SSD setup but it degraded quickly. It is detached but in the case. It was throwing significant errors and I decided to replace it..... that... would be a pretty big inflection point. What would you recommend? Note that since the new cache setup - It is still having issues saving VM stats - for instance a windows 10 VM base install settings wont stick.

Disable Docker and VM Manager in Settings

Run mover to see if you can get appdata and system moved to cache

Post new diagnostics

21 minutes ago, trurl said:

Disable Docker and VM Manager in Settings

Run mover to see if you can get appdata and system moved to cache

Post new diagnostics

 

Mover ran rather quickly - so I kicked it off twice. VM / Docker offline.

fractal-diagnostics-20211126-2209.zip

appdata still has files on the array. Mover won't move duplicates so that might be why.

 

What do you get from the command line with these?

du -h -d 1 /mnt/cache/appdata

du -h -d 1 /mnt/disk2/appdata

 

6 minutes ago, trurl said:

du -h -d 1 /mnt/cache/appdata

du -h -d 1 /mnt/disk2/appdata

 

 

root@Fractal:~# du -h -d 1 /mnt/cache/appdata
90M     /mnt/cache/appdata/Plex-Media-Server
756K    /mnt/cache/appdata/jellyfin
97M     /mnt/cache/appdata/pihole-dot-doh
520K    /mnt/cache/appdata/HandBrake
16M     /mnt/cache/appdata/jupyterlab
156K    /mnt/cache/appdata/MakeMKV
203M    /mnt/cache/appdata

 

root@Fractal:~# du -h -d 1 /mnt/disk2/appdata
0       /mnt/disk2/appdata/MakeMKV
0       /mnt/disk2/appdata
root@Fractal:~#

Looks like only one empty folder on disk2 appdata. You can delete with

rm -r /mnt/disk2/appdata

 

Executed. Appreciate the help working through this.

 

Post command: root@Fractal:~# du -h -d 1 /mnt/disk2/appdata
du: cannot access '/mnt/disk2/appdata': No such file or directory

 

If I were able to hook up the 120SSD externally - would it be worth trying to migrate data even as it is degrading?

 

If you have a port you can hook it up internally. Either way might be worth a look to see if anything on it would be useful.

root@Fractal:~# du -h -d 1 /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G
3.9G    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata
4.3M    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/photography
27G     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/domains
3.7M    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/system
1.2G    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/timemachine
8.1G    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/DataScience
2.9G    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/isos
43G     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G

root@Fractal:~# du -h -d 1 /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata
63M     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/pihole-dot-doh
3.2G    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/Plex-Media-Server
0       /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/deduper
16M     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/jupyterlab
0       /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/Plex-Meta-Manager
116K    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/portainer
712K    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/HandBrake
168K    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/MakeMKV
645M    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/jellyfin
68K     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/macinabox
20K     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/hoobs
24K     /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata/homebridge
3.9G    /mnt/disks/d2_Quadra_v3C_OCZ-562Q08GJ9QX3UU0G/appdata

1 hour ago, trurl said:

If you have a port you can hook it up internally. Either way might be worth a look to see if anything on it would be useful.

 

In terms of the apps - they have all been messed up since and nothing is kosher. Is this something that it makes sense to just overwrite select areas of the existing cache with the prior one or what would the best plan of attack be?

As you can see, each of those subfolder names are also user share names. User shares are just the top level folders on array and pools. So, the contents of those folders are those files from each of those user shares that were on that cache drive when you removed it.

 

It would have been better if you had transferred all that to the array and then some of it back to the new cache, and if you had asked before doing anything we would have recommended that and helped with it.

 

Now you will have to decide whether there are files on the new cache you want to keep, and which of the files from the old cache you want on the new cache and which on the array.

 

isos is the installation files for your VMs, such as the downloaded Windows install files.

 

appdata is the working storage of each of your docker containers, as you can see from the contents of appdata on that disk.

 

domains is the vdisks of your VMs.

 

system has docker.img, the executables of each of your docker containers, and libvirt.img, the xml of your VM configurations. If you haven't already tried to reinstall dockers and VMs those might be useful. If you have already tried to reinstall dockers and VMs it might be difficult to reuse those old ones.

 

All those others are files from other user shares that were still on cache, perhaps you were caching them and mover hadn't moved them to the array before you pulled the drive.

 

Do you know how to work with files directly on your server?

 

 

6 minutes ago, trurl said:

As you can see, each of those subfolder names are also user share names. User shares are just the top level folders on array and pools. So, the contents of those folders are those files from each of those user shares that were on that cache drive when you removed it.

 

It would have been better if you had transferred all that to the array and then some of it back to the new cache, and if you had asked before doing anything we would have recommended that and helped with it.

 

Now you will have to decide whether there are files on the new cache you want to keep, and which of the files from the old cache you want on the new cache and which on the array.

 

isos is the installation files for your VMs, such as the downloaded Windows install files.

 

appdata is the working storage of each of your docker containers, as you can see from the contents of appdata on that disk.

 

domains is the vdisks of your VMs.

 

system has docker.img, the executables of each of your docker containers, and libvirt.img, the xml of your VM configurations. If you haven't already tried to reinstall dockers and VMs those might be useful. If you have already tried to reinstall dockers and VMs it might be difficult to reuse those old ones.

 

All those others are files from other user shares that were still on cache, perhaps you were caching them and mover hadn't moved them to the array before you pulled the drive.

 

Do you know how to work with files directly on your server?

 

 

 

I will definitely make a note of more research / posting if needed before... something like this again..

 

In terms of working with files directly - basic linux moving /copying commands - yes and permissions, but not sure much beyond that. Is this process going forward similar to a git merge error where it needs to be parsed out and reviewed? Appreciate the thorough explanation of the main directory components

 

1 minute ago, unraidable said:

basic linux moving /copying commands

Midnight Commander (mc from the command line) is builtin and how I usually do it.

 

docker.img is often not important since those executables are downloaded again simply by reinstalling the containers. Previous Apps on the Apps page will use the templates from the previous installs, but if you have already reinstalled some of those then your reinstallations have already become the previous install.

 

Since Docker and VM Manager are disabled, I guess you could just get rid of appdata, domains, system on cache and copy those from that disk to cache and see what happens when you enable them.

I moved it over but when I fired them up the dockers were empty and the VMs were not working. I started to rebuild the docker and settings and such are sticking - which is awesome. I am still experiencing some connectivity issues where the unraid management portal disappears and is unreachable for a few minutes at a time.