[systemd-devel] How do I easily resolve conflicting jobs in custom units?

[John Florian]

I build an mostly-stateless appliance OS derived from Fedora (25 ATM)
and have several custom units to make it all possible.  My units had
worked great with F21, but are now giving me problems with F25.  One
pair of the custom units do some trickery to relocate sshd host keys
from /etc/ssh to an alternate location that provides persistence:

$ systemctl cat sshd-persist-keys.service
# /usr/lib/systemd/system/sshd-persist-keys.service
[Unit]
Description=OpenSSH server - persist volatile keys for the AOS
After=sshd-keygen.target
Before=sshd.service
Wants=sshd-keygen.target

[Service]
ExecStart=/usr/sbin/sshd-persist-keys
Type=oneshot
RemainAfterExit=yes

[Install]
WantedBy=multi-user.target


$ systemctl cat sshd-restore-keys.service
# /usr/lib/systemd/system/sshd-restore-keys.service
[Unit]
Description=OpenSSH server - restore persisted keys for the AOS
After=aos-storage-init.service
Before=sshd-keygen at rsa.service sshd-keygen at ecdsa.service sshd-keygen at ed
25519.service

[Service]
ExecStart=/usr/sbin/sshd-restore-keys
Type=oneshot
RemainAfterExit=yes

[Install]
WantedBy=multi-user.target


I found that on some boots, sshd wasn't getting started.  With the help
of booting with systemd.log_level=debug I learned:

$ sudo journalctl | grep conflict
Mar 22 16:11:42 localhost systemd[1]: sshd.service: Looking at job
sshd.service/start conflicted_by=no
Mar 22 16:11:42 localhost systemd[1]: sshd.service: Looking at job
sshd.service/stop conflicted_by=yes
Mar 22 16:11:42 localhost systemd[1]: sshd.service: Fixing conflicting
jobs sshd.service/start,sshd.service/stop by deleting job
sshd.service/start
Mar 22 16:11:42 localhost systemd[1]: sshd.socket: Looking at job
sshd.socket/stop conflicted_by=no
Mar 22 16:11:42 localhost systemd[1]: sshd.socket: Looking at job
sshd.socket/start conflicted_by=no
Mar 22 16:11:42 localhost systemd[1]: sshd.socket: Fixing conflicting
jobs sshd.socket/stop,sshd.socket/start by deleting job
sshd.socket/stop
Mar 22 16:11:42 localhost systemd[1]: sshd-keygen.target: Looking at
job sshd-keygen.target/start conflicted_by=no
Mar 22 16:11:42 localhost systemd[1]: sshd-keygen.target: Looking at
job sshd-keygen.target/stop conflicted_by=no
Mar 22 16:11:42 localhost systemd[1]: sshd-keygen.target: Fixing
conflicting jobs sshd-keygen.target/start,sshd-keygen.target/stop by
deleting job sshd-keygen.target/stop


I'm sure systemd-analyze is my friend here, but I've not found an
effective way to grok the problem.  s-a dot either gives me too much to
comprehend or not enough to make evident the problem.  I've tried
feeding it different PATTERNS like: none at all, sshd.service, ssh-
keygen.target, sshd-{persist,restore}-keys.service and more.  I either
see a whole forest or just a few trees that seem to be getting along
happily together.

What advice can you suggest?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/systemd-devel/attachments/20170322/f6a226ae/attachment.html>