On Sat, 2024-09-21 at 13:16 +0100, Patrick O'Callaghan wrote:
> [Rather than adding to an existing thread (which discusses AVCs from
> crontab among other things, this is starting from a clean slate to
> reduce confusion.]
>
> 1. I run httpd via systemd for a small personal web service. This
> works.
>
> 2. I suspend my system overnight and wake it automatically in the
> morning). This turns out to be irrelevant, because the problem reported
> here occurs even when I do it manually.
>
> 3. On resuming, everything is up *except* httpd.
>
> 4. The journal (see below) shows that httpd does resume with the rest
> of the system, but then is immediately deactivated. The reason for this
> is not stated (there's no error message), but may be because on the
> system resume the network is not yet up. A similar thing seems to
> happen to certbot, though I haven't looked at it in detail.
>
> Here's the relevant section from the journal:
>
> Sep 21 13:04:12 Bree systemd[1]: httpd.service: Deactivated successfully.
> Sep 21 13:04:12 Bree systemd[1]: httpd.service: Consumed 2.980s CPU time.
> Sep 21 13:04:14 Bree NetworkManager[1214]: <info> [1726920254.0080] device
> (enp12s0): carrier: link connected
> Sep 21 13:04:14 Bree kernel: r8169 0000:0c:00.0 enp12s0: Link is Up -
> 1Gbps/Full - flow control rx/tx
>
> I assume there's a systemd mechanism for getting httpd to wait for the
> network before resuming, but I've no idea how to accomplish this.
>
> poc
I made some simplifications and the problem seems to have gone away, at
least in the past few days, though I'm not entirely sure what made the
difference. Fingers crossed.
1) I'm no longer doing anything with crontab
2) I've replaced apachectl with systemctl httpd. I don't see any reason
why this should matter, but there you are.
3) I'm not trying to be "graceful", just stopping/starting httpd.
4) Although I did test the httpd.socket service, I'm not using it.
5) I have the following script in /lib/systemd/system-
sleep/apache.sleep:
#!/bin/bash
# Hooks to suspend/resume Apache
#
LOG=/var/log/httpd/my-log
log() {
(echo "$(date): Apache $1") >> $LOG
}
running() {
systemctl is-active httpd > /dev/null
return $?
}
case $1 in
pre) if running; then
if systemctl stop httpd; then
log "stopping"
else
log "stop failed"
fi
else
log "not running"
fi
;;
post) if running; then
log "resumed"
else
log "starting ..."
if systemctl start httpd; then
log "... succeeded"
else
log "... failed"
fi
fi
esac
Thanks to all who chipped in.
poc
--
_______________________________________________
users mailing list -- users@lists.fedoraproject.org
To unsubscribe send an email to users-le...@lists.fedoraproject.org
Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam, report it:
https://pagure.io/fedora-infrastructure/new_issue
