Lost my VMs after adding a cache drive!

[JasenHicks]

So I may have done something foolish during setup of my new unRAID server. I initially set it up, added my 2x SSD cache drives and my 3x 8TB Iron Wolf drives to the array and my 4th 8TB Iron Wolf as a Parity drive. Cool, right? Yup. Had it all up and it was great until I started copying files from one "nas" to this and it chewed up my 10GBps connection and right through my 500GB of SSD Cache (2x in RAID 0 by default). I thought this seems silly, why not take the advice of some others on the forums here (since I had been lurking about) and "remove" the parity drive and cache drive to write everything direct to the array and then add parity and cache back later? So onward I went, stopped the array, removed the cache drives and parity and started the copy again. Nice.

 

While it was copying, I was dorking about with VMs and setup a Windows 10 VM that was running dropbox to sync it all onto one of the shares in the VM (using some of the necessary trick f-ery since Dropbox doesn't allow network targets as the Dropbox folder). In my mind this was good so that I could keep that one drive synced automatically with the interweb cloud AND have access to it via my network from my multitude of PCs without having to run dropbox on each of them; I mean why have this insanely overkill dual Xeon server and 160GB of ECC RAM if you aren't going to do something like this anyways? Plus I couldn't follow Space Invaders tutorial and have it work since I am running the newer unRAID and things didn't perfectly jive... I may just not be smart enough or have enough experience yet to use rclone. I digress.

 

This morning the copy was done, everything was great. So I stopped the array, added the cache drives back, added the parity drive back, and off to work I went while I racked up an expensive electricity bill letting the server do its thing. When I got home, I finished the unRaid tutorial on adding cache and such back and voila, everything was good? NOPE! My VMs are nowhere to be found. I stopped DOCKER and VM Manager, verified the SYSTEM and APPDATA and DOMIANS were set to "cache preferred" and clicked the MOVE button. Let it finish. Still no VMs. I checked the paths for VMs and its all properly (I think) pointing to the cache drives. See picture.

 

In VMs nothing is listed and I am at a loss on what to do. I checked the cache drives and I see that it has both vdisk1.img and vdisk2.img (coincidentally its on the array still too, but I am afraid to delete anything).

 

Any advice? I am only a couple days into my trial and want to throw wads of cash at LimeTech, but want to make sure this is all going to work how I envision it. Appreciate the support and hopefully the warm welcome... no "idiot pitchforks" please. Pretty please?

[JorgeB]

You likely have duplicate VM and docker images from when you worked on them without cache, post the output of:

 

find /mnt -name libvirt.img

 

[JasenHicks]

I think I know where you going... but I patiently await your reply so I don't guarantee I jack this up

[JorgeB]

Stop the VM service, change the libvirt storage location to /mnt/disk1/system/libvirt/libvirt.img, start service and your VM should be back.

 

You can then delete the libvirt.img on cache and move the other one there with the mover (with the VM service stopped and will need to adjust path in the end).

[JasenHicks]

Whelp... it did something. Not what I expected though

 

When I did stop it and change the storage location, it did show my Win10 VM.

[JasenHicks]

I'm a jackwagon.... turns out you need to have the libvirt.img file in the storage location as well. Thanks for the assist! ::BeerForYou::

[Tukke]

have you fixed it?

[Squid]

34 minutes ago, Tukke said:

have you fixed it?

If you're having problems, it would be better to create your own thread with the details and your diagnostics rather than replying to a 3+ year old thread that is solved.