Also worth noting that we didn’t change qemu between LXD 4.2 and 4.3, nor have we changed anything that relates to QMP or VM status, so a downgrade to 4.2 almost certainly wouldn’t help there.
Yes I think it is just a coincidence this problem happened the same day as the 4.3 upgrade.
By the way I love the new “lxc start --console vm” - no longer do you need lightning fast reflexes! You can view the entire boot up process. Very useful. Thanks
Yeah, I added that flag after having to do lxc start NAME ; lxc console NAME
way too many times
One thing that may be worth a shot just in case is:
- lxc config set PiDebian security.secureboot true
- lxc config set PiDebian security.secureboot false
This will force a re-generation of the UEFI NVRAM for the VM.
Normally the NVRAM wouldn’t prevent qemu from booting, but I remember arm64 being a bit picky about exact size of that NVRAM so maybe there’s something in there that’s confusing qemu.
Though that really doesn’t explain things working fine if started in a different order…
New error:-
wolfgang@PiLXD:~ $ lxc config set PiDebian security.secureboot true
wolfgang@PiLXD:~ $ lxc config set PiDebian security.secureboot false
wolfgang@PiLXD:~ $ lxc start PiDebian
Error: write /var/snap/lxd/common/lxd/virtual-machines/PiDebian/config/lxd-agent: no space left on device
Try lxc info --show-log PiDebian
for more info
lxc info --show-log PiDebian
Name: PiDebian
Location: none
Remote: unix://
Architecture: aarch64
Created: 2020/07/02 17:49 UTC
Status: Stopped
Type: virtual-machine
Profiles: default
Pid: 2473
Resources:
Processes: 0
Snapshots:
BeforeDocker (taken at 2020/07/03 17:04 UTC) (stateless)
Log:
qemu-system-aarch64: terminating on signal 15 from pid 1 ()
df -h
Filesystem Size Used Avail Use% Mounted on
udev 1.9G 0 1.9G 0% /dev
tmpfs 380M 3.9M 376M 2% /run
/dev/mmcblk0p2 59G 46G 11G 83% /
tmpfs 1.9G 0 1.9G 0% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 1.9G 0 1.9G 0% /sys/fs/cgroup
/dev/loop0 49M 49M 0 100% /snap/core18/1708
/dev/loop1 64M 64M 0 100% /snap/lxd/15906
/dev/loop3 64M 64M 0 100% /snap/lxd/15938
/dev/loop2 49M 49M 0 100% /snap/core18/1753
/dev/loop4 26M 26M 0 100% /snap/snapd/8147
/dev/loop5 27M 27M 0 100% /snap/snapd/7780
/dev/mmcblk0p1 253M 98M 155M 39% /boot/firmware
tmpfs 380M 0 380M 0% /run/user/1001
tmpfs 1.0M 0 1.0M 0% /var/snap/lxd/common/ns
Problem solved: Even though I had 11 GB free on my SD card, the BTRFS partition that I initially setup when I ran lxd init, was full! So all the sluggish performance and everything else was because of this. I deleted all used images and freed up space and all is good now Sorry for the trouble…