#813334 colord: syslog spam after failure to get session

Package:
colord
Source:
colord
Description:
system service to manage device colour profiles -- system daemon
Submitter:
Dominique Brazziel
Date:
2020-11-13 13:39:03 UTC
Severity:
important
#813334#5
Date:
2016-01-31 18:30:08 UTC
From:
To:
    Every time the CUPS scheduler wakes up and runs, colord
drops 3 messages in the message log:


Jan 30 09:02:02 dell2800 colord[587]: (colord:587): Cd-WARNING **: failed to get session [pid 485]: No such device or address
Jan 30 09:02:02 dell2800 colord[587]: (colord:587): Cd-WARNING **: failed to get session [pid 485]: No such device or address
Jan 30 09:02:02 dell2800 colord[587]: (colord:587): Cd-WARNING **: failed to get session [pid 485]: No such device or address
Jan 30 09:07:42 dell2800 colord[587]: (colord:587): Cd-WARNING **: failed to get session [pid 4198]: No such device or address
Jan 30 09:07:42 dell2800 colord[587]: (colord:587): Cd-WARNING **: failed to get session [pid 4198]: No such device or address
Jan 30 09:07:42 dell2800 colord[587]: (colord:587): Cd-WARNING **: failed to get session [pid 4198]: No such device or address

I'm guessing there is a message for each CUPS printer queue as I have 3
defined.  This has become more of a problem because it seems like the
CUPS mechanism to keep cupsd alive forever is broken, but that's another
issue.
#813334#10
Date:
2016-03-16 13:36:41 UTC
From:
To:
Exactly how is this bug distinct from #750533? (Why shouldn't they be merged?)

I'm seeing similar[*] warnings from colord, which I traced down to the fact
that cupsd does not belong to a systemd session. From its /proc/self/cgroup:

1:name=systemd:/system.slice/cups.service

I don't see anything wrong with that state of affairs (cupsd is a system-wide
service, not associated with any particular seat), so I don't think this needs
to be logged at severity WARNING (unless it's somehow a mistake for a process
not associated with a seat to send messages to colord, but I don't see why that
should be the case). Can we get that g_warning() changed to g_debug()?

(On a not entirely unrelated note, I don't think this bug deserves to be at
severity important.)

[*] I'm seeing ENOENT (as in #750533) rather than ENODEV, but this is with
jessie's systemd and kernel.
#813334#17
Date:
2017-01-15 07:23:17 UTC
From:
To:
I also see this on a system running stretch

It is important to stamp out spam in Syslog

Many people use automatic log monitoring tools, e.g.
syslog-nagios-bridge[1], to alert on errors in their logs.  They only
want to be alerted when something genuinely needs their attention.  Or
when looking at a report in LogAnalyzer[2], it can be hard to spot the
real problems if there are lots of these more trivial messages.

When a message is logged with this severity, it would also be useful if
the message contained some type of call to action or link to further
information (e.g. "please check X")

Regards,

Daniel



1. https://packages.qa.debian.org/s/syslog-nagios-bridge.html
2. https://packages.qa.debian.org/l/loganalyzer.html
#813334#22
Date:
2020-11-13 13:36:02 UTC
From:
To:
Hello,
Hope you and your family is doing well and safe.

I am just following up to see if you are interested in acquiring the Attendees/Visitors  list of,

SACO Students Fair
26 - 28 Nov 2020
Stockholm International Fairs, Stockholm, Sweden
Count=2143

Each record of the list contains Contact Name, Email Address, Phone No, Title, Company Name, URL/Website, City, Country, Zip code.

Let me know your thoughts so that we can send you the cost and additional information.


Best Regards,
Jennifer Leoni
Business Executive