VMs pause imediately after startup

[bubbleman441]

I'm having a rough time trying to solve my issues which is a two part problem and I desperately need help so I can access my VMs.

 

1) My VMs are all pausing as soon as they boot to the operating system. Every once in awhile they will load the desktop. VMs are in domains on the cache. According to the main page I've used 558GB out of 750GB. Am I creating instability by using too much disk space?

 

 

 

2) I had previously had my VIDEOS and PICTURES shares selected to "yes" for writing to cache. I then changed them both to "no" thinking that all files would be transferred to the array once the mover started. I guess this is not so. How do I go about moving what's on the cache to the array so I can free up space? I can't seem to locate my cache drive using Krusader. Nothing shows up under the /mnt directory. To add to the confusion, after taking a look at my shares on the cache, they don't add up to 558GB (3.02GB + 283GB + 1.29GB + 3.95GB = 291.26GB). Please help me understand what's going on here. What is the best course of action?

Thanks in advanced for the help.

 

Edited July 27, 2019 by bubbleman441

[itimpi]

If you want mover to move files from cache to array then a value of “Yes” is the only one that will achieve this.   A value of “No” means do not put new files onto the cache but leave any that are already there alone.

 

i notice that you have two disks in the cache of different sizes.    Are they running in the default mode of RAID1 or have you taken manual action to put them into a different mode?   
 

you should post your system diagnostics zip file (obtained via Tools->Diagnostics) so you can get some more informed feedback.

[JorgeB]

2 hours ago, bubbleman441 said:

My VMs are all pausing as soon as they boot to the operating system.

This almost certainly means they are out of space, diags will confirm.

[bubbleman441]

8 hours ago, itimpi said:

If you want mover to move files from cache to array then a value of “Yes” is the only one that will achieve this.   A value of “No” means do not put new files onto the cache but leave any that are already there alone.

 

i notice that you have two disks in the cache of different sizes.    Are they running in the default mode of RAID1 or have you taken manual action to put them into a different mode?   
 

you should post your system diagnostics zip file (obtained via Tools->Diagnostics) so you can get some more informed feedback.

I just followed Space Invaders tutorial on making a cache pool.. Can different sized disks in a cache pool create issues? Diagnostics attached.

 

tower-diagnostics-20190727-1531.zip

[itimpi]

19 minutes ago, bubbleman441 said:

I just followed Space Invaders tutorial on making a cache pool.. Can different sized disks in a cache pool create issues? Diagnostics attached.

Only when it leads to confusion about usable space as the GUI can be misleading 

 

If using the default RAID1 profile then with two disks of unequal size only space equivalent to the smaller drive is available.  If you have more than 2 drives in the pool then things get more complicated.  There is a web page you can use to see how the number of disks and BTRFS RAID profile affect each other as far as available space is concerned.

 

Another potential source of confusion is that VM vdisks are by default created as 'sparse' files which means they will occupy less physical space on the drive than the initial allocation, but can grow their physical requirements up to that allocated size as they are used.

[JorgeB]

Like suspect you are out of space, also because of a bug with current release your pool is not redundant, after freeing some space convert metadata to raid1:

 

btrfs balance start -dconvert=raid1 -mconvert=raid1 /mnt/cache

Edit: data also needs to be converted since some part is also single profile.

Edited July 29, 2019 by johnnie.black

[bubbleman441]

On 7/27/2019 at 9:09 AM, itimpi said:

Only when it leads to confusion about usable space as the GUI can be misleading 

 

If using the default RAID1 profile then with two disks of unequal size only space equivalent to the smaller drive is available.  If you have more than 2 drives in the pool then things get more complicated.  There is a web page you can use to see how the number of disks and BTRFS RAID profile affect each other as far as available space is concerned.

 

Another potential source of confusion is that VM vdisks are by default created as 'sparse' files which means they will occupy less physical space on the drive than the initial allocation, but can grow their physical requirements up to that allocated size as they are used.

Thanks for clearing this up. I realized my cache pool was in RAID1. I was hoping to have a full 750GB of usable data with no redundancy. I guess this isn't possible with RAID0.

[JonathanM]

2 hours ago, bubbleman441 said:

I guess this isn't possible with RAID0.

Not possible. You want single, not RAID0. Plug your numbers in here and change the RAID levels to see what the usable capacity will be.

https://carfax.org.uk/btrfs-usage/

[bubbleman441]

2 hours ago, jonathanm said:

Not possible. You want single, not RAID0. Plug your numbers in here and change the RAID levels to see what the usable capacity will be.

https://carfax.org.uk/btrfs-usage/

Sorry for the newb questions. So I cannot make a cache pool with a 250GB and 500GB SSD giving me 750 usable disk space? I'm assuming "single" means one single disk?

Edited August 3, 2019 by bubbleman441

[JorgeB]

2 hours ago, bubbleman441 said:

Sorry for the newb questions. So I cannot make a cache pool with a 250GB and 500GB SSD giving me 750 usable disk space? I'm assuming "single" means one single disk?

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=480421