You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
ostree has nice flexible support for injecting metadata into commits. Cool. We've been extensively using it in various ways, e.g coreos-assembler injects basearch and rpm-ostree injects the initramfs-args.
cosa build-fast omits all that stuff. One noticeable side effect of missing the basearch is that zincati goes into a start-die-restart loop (should probably fix cosa to have "this is a fastbuild" flag or something and zincati gracefully shuts down if so?). The only real harm is the spam in the journal, but still.
And a side effect of missing the initramfs-args is that testing rpm-ostree changes in a cosa build-fast and then testing replacing the kernel in there will end up generating a broken initramfs (this took me longer than it should have to figure out, I got distracted by thinking it was selinux because for some reason we're getting a new denial in 33 when rebuilding the initramfs client side).
(I use cosa build-fast a lot)
The text was updated successfully, but these errors were encountered:
This is still relevant I think, though specifically in the initramfs case, I think that was fixed with the recent changes in rpm-ostree to use the treefile instead IIRC.
ostree has nice flexible support for injecting metadata into commits. Cool. We've been extensively using it in various ways, e.g coreos-assembler injects
basearch
and rpm-ostree injects theinitramfs-args
.cosa build-fast
omits all that stuff. One noticeable side effect of missing the basearch is that zincati goes into a start-die-restart loop (should probably fix cosa to have "this is a fastbuild" flag or something and zincati gracefully shuts down if so?). The only real harm is the spam in the journal, but still.And a side effect of missing the
initramfs-args
is that testing rpm-ostree changes in acosa build-fast
and then testing replacing the kernel in there will end up generating a broken initramfs (this took me longer than it should have to figure out, I got distracted by thinking it was selinux because for some reason we're getting a new denial in 33 when rebuilding the initramfs client side).(I use
cosa build-fast
a lot)The text was updated successfully, but these errors were encountered: