Aug 13 22:38:27 bkor spamd[1912]: Aug 13 22:38:27.856 [1912] warn: server socket setup failed, retry 9: spamd: directory for /var/run/spamd/spamd.socket does not exist, exiting Aug 13 22:38:28 bkor spamd[1912]: Aug 13 22:38:28.857 [1912] error: spamd: directory for /var/run/spamd/spamd.socket does not exist, exiting Aug 13 22:38:28 bkor spamd[1912]: spamd: directory for /var/run/spamd/spamd.socket does not exist, exiting Aug 13 22:38:28 bkor systemd[1]: spamassassin.service: control process exited, code=exited status=255 Aug 13 22:38:28 bkor systemd[1]: Unit spamassassin.service entered failed state. lack of setting up /var/run/spamd
Assignee: bugsquad => remcoTarget Milestone: --- => Mageia 3
Status: NEW => ASSIGNED
CC: (none) => mageia
@colin I see the same as Olav does on my Cauldron box. However, in the .spec file for this package I do have install -d %{buildroot}/var/run/spamd Furthermore, this spamd directory does show up under /var/run.runmove~/, but not under /var/run/ . Could this in any way be related to the big /usr move?
Yes it's 100% usrmove related. It needs tmpfiles.d now. I've been lazy so have not fixed up all the packages yet. I'll hopefully do a "spurt" soon.
@coling, Thanks for that. I had not seen this package listed as needing fixing up. Is this anything simple I can do? If it was on the mailinglist, a message ID on how / what to fix will hopefully suffice for me to fix this (and perhaps lend a hand with other packages needing a fix). :-)
Should be fixed now (not tested just updated and submitted so please let me know). Feel free to check the commit. It's basically a matter of adding a tmpfiles.d file which ensures folders etc. are correctly created on fresh boot. Some suggested improvements: 1. Make spamd socket activated such that it can be started automatically when something needs it (assuming this is supported internally - might need some modifications to accept the socket - worth checking fedora). 2. Why is spamd run as root? Seems like a bad idea to me. We should run it as a less privileged user.
Thanks Colin, fix confirmed. I'll look at your suggestions later. As for #2, I guess this is because it wants to bind to port 783 (privileged port) by default.
Status: ASSIGNED => RESOLVEDResolution: (none) => FIXED