That's a REALLY bad resolution from our end. We rely on firewalld; I cannot "simply remove it" to test.
I suspect I also can't get rid of cloud-init though, so that's quite the pickle. I suppose it's time to learn nftables ... or dive in and try and fix firewalld. Thanks for letting me know; from a Debian perspective this issue is likely a noop. - Wyatt On 8/5/25 14:14, Noah Meyerhans wrote: >> We encountered an issue where some core services were unexpected >> unavailable following a reboot (most notably clamav-daemon; fortunately >> the software relying upon it "failed closed" so no uploads went unscanned). >> >> Investigating this, there are several packages involved and I'm >> uncertain about the particular service file so apologies if this is >> directed towards the wrong package. >> >> Aug 01 15:00:11 hostname systemd[1]: firewalld.service: Job >> cloud-init.service/start deleted to break ordering cycle starting with >> firewalld.service/start > I believe this is the same issue as > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1025618, which is > fixed in trixie. Unfortunately it's fixed by adding a Conflicts > relationship with firewalld. > > If you remove firewalld, is the cyclic dependency resolved? > > noah >
