David Edler
David Edler
> Why not both? :) Show disk usage first (The sum of all partitions. This will work for both scenarios of a custom partition layout and the default image layout)...
> Please either use relative URLs, allowing serving from any directory (though requires matching logic from end), or add configurable path prefix option. Thank you for your request, I understand...
> Any news on this? This is on our current roadmap to be finished until April 2026. I think it will land after the upcoming LXD 6.6 but in time...
I think that is not true, the start operation fails ``` edlerd@EDLERD-T14:~$ lxc start robust-haddock ; lxc operation list ^[[AError: Failed to run: /snap/lxd/current/bin/lxd forkstart robust-haddock /var/snap/lxd/common/lxd/containers /var/snap/lxd/common/lxd/logs/robust-haddock/lxc.conf: exit status...
I think it might be better to update the status of the instance at the end of an operation. Contrary to the current implementation that is changing the status when...
Thanks for the code pointers, I reproduced this with debug logging. See below for details, I marked the lock lines with a `>`. I think the operation lock works while...
> With the reality that it did start and got far enough to trigger the stop hook when it did finally stop: > > ``` > DEBUG [2024-09-05T12:07:34+02:00] Handling API...
I see, thanks for the explanation. Now I get why it is so tricky with the lower level barrier in between. Another idea: We introduce a global list of instances...
After contemplating on this for half a day: As far as the LXD-UI is concerned, there is a much simpler solution. We can defer the clearing of the cache to...