Bug 1784884

Summary: cups.service doesn't execute automatically on request
Product: Red Hat Enterprise Linux 8 Reporter: Renaud Métrich <rmetrich>
Component: cupsAssignee: Zdenek Dohnal <zdohnal>
Status: CLOSED ERRATA QA Contact: Petr Dancak <pdancak>
Severity: medium Docs Contact:
Priority: medium    
Version: 8.1CC: msekleta, pdancak
Target Milestone: rcKeywords: Patch, TestCaseProvided, Triaged
Target Release: 8.0   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: cups-2.2.6-34.el8 Doc Type: No Doc Update
Doc Text:
Story Points: ---
Clone Of:
: 1822154 (view as bug list) Environment:
Last Closed: 2020-11-04 01:42:20 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1817044, 1822154, 1825061    

Description Renaud Métrich 2019-12-18 15:11:23 UTC
Description of problem:

The cups.service is an on-demand service, usually triggered by cups.path or cups.socket (unix socket) systemd units.
In order to configure cups, it also needs to be triggered when connecting to administration port (631).

This just works fine when the graphical interface is used and a user logs in. Indeed, in such case, the first time this happens (usually it's through post-configuration during first boot after installation), /usr/libexec/gsd-print-notifications executes, triggering a connection on /var/run/cups/cups.sock, which itself starts cups.service, which finally creates /var/cache/cups/org.cups.cupsd file.
Upon next boot, /var/cache/cups/org.cups.cupsd exists, so cups.path triggers cups.service unit automatically.

Unfortunately, this workflow doesn't work when there is no graphical user interface: nothing triggers /var/run/cups/cups.sock, no cups.service never starts. Upon reboot, since there is no /var/cache/cups/org.cups.cupsd, cups.service doesn't start either.

To solve this, in my opinion the cups.socket service should also listen on the administration port (by default ::1:631 and 127.0.0.1:631).
That way, even when no graphical interface is used, the cups.service unit would execute on-demand.


Version-Release number of selected component (if applicable):

cups-2.2.6-28.el8.x86_64


How reproducible:

Always


Steps to Reproduce:
1. Install a system with cups packages but no graphical interface

Actual results:

cups.service never started automically since there is no way to trigger it.

Additional info:

A simple fix is to modify the cups.socket unit to listen on the administration port, as shown with the following drop-in:

-------- 8< ---------------- 8< ---------------- 8< ---------------- 8< --------
# mkdir /etc/systemd/system/cups.socket.d
# cat > /etc/systemd/system/cups.socket.d/listen_admin_port.conf << EOF
[Socket]
ListenStream=::1:631
ListenStream=127.0.0.1:631
EOF

# systemctl daemon-reload
# systemctl restart cups.socket
-------- 8< ---------------- 8< ---------------- 8< ---------------- 8< --------

The administrator may then relax the permissions (usually done in /etc/cups/cupsd.conf through "Listen" stanza) by altering the cups.socket unit file.

Comment 19 errata-xmlrpc 2020-11-04 01:42:20 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Low: cups security and bug fix update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2020:4469