Bug 1073481

This request was not resolved in time for the current release.
Red Hat invites you to ask your support representative to
propose this request, if still desired, for consideration in
the next release of Red Hat Enterprise Linux.

Problem here is that there is difference between unit file and unit. is-active and is-failed operates strictly on units. We can't say that pacemaker is inactive, because we simply just see that there is no such unit and we can't decide if it is the case that service was stopped and removed by garbage collector or there is no unit file at all.
So I am sorry but this is WONTFIX/CANTFIX/NOTABUG for me.

Lukas, I don't understand your explanation here.

What's the difference between unit and unit-file? systemd can clearly determine if the service a) exists and b) is active or stopped or dead or failed.

So why can't this knowledge get propagated via is-failed command?
# systemctl status pacemaker
pacemaker.service - Pacemaker High Availability Cluster Manager
   Loaded: loaded (/usr/lib/systemd/system/pacemaker.service; disabled)
   Active: inactive (dead)

# systemctl is-failed pacemaker
unknown

This does not make much sense to me.

Unit file is file on the disk, unit is a structure in systemd memory, unit does not need an unit file and to save memory we don't have units for unit files which are not needed.

ok and now once again (as you did not respond to it):

So why can't this knowledge get propagated via is-failed command?
# systemctl status pacemaker
pacemaker.service - Pacemaker High Availability Cluster Manager
   Loaded: loaded (/usr/lib/systemd/system/pacemaker.service; disabled)
   Active: inactive (dead)

# systemctl is-failed pacemaker
unknown

This actually seems to work on F21:

[adamw@adam gnome-initial-setup (f21 %)]$ systemctl status iscsid.service
● iscsid.service - Open-iSCSI
   Loaded: loaded (/usr/lib/systemd/system/iscsid.service; disabled)
   Active: inactive (dead)
     Docs: man:iscsid(8)
           man:iscsiadm(8)
[adamw@adam gnome-initial-setup (f21 %)]$ systemctl is-failed iscsid
inactive
[adamw@adam gnome-initial-setup (f21 %)]$ 

did something change, Lukas?

The same thing works in rhel

[lnykryn@notas initscripts(rhel7-branch)]$ systemctl is-active avahi-daemon.service
unknown
[lnykryn@notas initscripts(rhel7-branch)]$ systemctl status avahi-daemon.service
avahi-daemon.service - Avahi mDNS/DNS-SD Stack
   Loaded: loaded (/usr/lib/systemd/system/avahi-daemon.service; disabled)
   Active: inactive (dead)

[lnykryn@notas initscripts(rhel7-branch)]$ systemctl is-active avahi-daemon.service
inactive


Status is more complex function with creates a unit in systemd manager. Than you can view its status through is-active until garbage collectors eats it.

ah, I see.

well, then it works in a very inconsistent way. The commands should return the same result unless the service state has changed. Making one's result depend on what has been run before is more kind of your internal magic than anything user friendly ;).

Is there any dependable machine-parsable command that would produce reliable output?

> Is there any dependable machine-parsable command that would produce reliable
> output?

Sure

$ systemctl show bluetooth.service -p ActiveState
ActiveState=inactive

And there is a stability promise for this command. So you should definitely use this one in scripts.

*** Bug 1417251 has been marked as a duplicate of this bug. ***