srd424

> Can you show a full `incus config show --expanded ostree3`? Oops, sorry, missed this. I assume you meant for the VM, not the (failed) virtiofs disk device: ``` architecture:...

> We don't restart workloads on upgrade, only the control plane (API) goes down during the upgrade. Oh, worth knowing, thanks! I did wonder if that was the case but...

BTW, can confirm this does happen on 6.0.1 too.

> For now the trick you can use is start the VM with a higher allocation than needed and the reduce limits.memory which will use the memory balloon driver to...

Just remembered, it needs some tidying up, doesn't need ramdisk support for example.

Hmmm, liminix kernels which work fine when tftpbooted fail to load to when written to flash. Best guess is the built-in uboot decompresses it to a different address in that...

(I have just noticed that this system uses nss-extrausers, I should be able to replace that with nss-systemd, which might be an interesting/useful experiment.) On Sun, 4 Aug 2024, 11:16...

I just hacked up your own GHA workflows! Nightly: https://github.com/srd424/nsncd/actions/runs/10235189099/job/28315666470#step:6:83 Stable: https://github.com/srd424/nsncd/actions/runs/10235156172/job/28315590851#step:6:254 Probably trivial to someone who understands the rust ecosystem, but I'm a bit of a dinosaur :(

FWIW, I've converted the user records that were using nss-extrausers into json for nss-systemd instead, and dropped extrausers from nsswitch.conf - 24 hours should be enough to tell if it's...

~36 hours without a crash, I'm going to tentatively call that fixed! My immediate need for a rebuild is therefore gone, feel free to close this, or alternatively leave it...