Yes you are right, I meant to say disk queuing.
-- Eric
On 02/10/11 23:29, Rainer Gerhards wrote:
-----Original Message-----
From: rsyslog-boun...@lists.adiscon.com [mailto:rsyslog-
boun...@lists.adiscon.com] On Behalf Of Eric Schoeller
Sent: Friday, February 11, 2011 1:48 AM
To: rsyslog-users
Subject: Re: [rsyslog] Queuing subsystem and message filtering
Okay,
I started up rsyslog with debugging twice. Once with queuing turned on,
once without.
What exactly do you mean with "queing turned on"? You mean disk queue
enabled, right? I am asking because rsyslog always queues, but via different
methods. I'd like to be sure to look at the right thing (and, yes, I could
extract that from the debug log as well but I'd like me to save that
time...).
Thanks,
Rainer
I immediately ran the test script that sends one message
to each facility at each severity and then killed the rsyslog daemon
...
so the results between the two *should* be fairly identical, there is
minimal additional syslog traffic on the machine at this time.
Immediately, I noticed:
host: /tmp> grep -i logserver rsyslog.queue.debug | wc -l
173
host: /tmp> grep -i logserver rsyslog.noqueue.debug | wc -l
142
As I suspected, more logs were being sent to the logserver with queuing
turned on (ie. they weren't getting filtered). I've posted the debug
logs here:
http://spark.colorado.edu/rsyslog/
I am somewhat new to rsyslog, and I've certainly never trolled through
a
debug file. I may have a chance to take a look tonight, but given my
familiarity with the software it may be difficult to zero in on
anything.
Thanks!
-- Eric
Rainer Gerhards wrote:
Could you create a debug log with such a message? It should tell us
what is
going on. Instructions:
http://www.rsyslog.com/doc/troubleshoot.html
Rainer
-----Original Message-----
From: rsyslog-boun...@lists.adiscon.com [mailto:rsyslog-
boun...@lists.adiscon.com] On Behalf Of Eric Schoeller
Sent: Thursday, February 10, 2011 7:11 PM
To: rsyslog-users
Subject: Re: [rsyslog] Queuing subsystem and message filtering
Oh, sorry my mail client must have munged those lines. I am getting
a
clean start-up and shutdown:
Feb 10 11:02:07 host rsyslogd: [origin software="rsyslogd"
swVersion="5.6.3" x-pid="15452" x-info="http://www.rsyslog.com";]
exiting
on signal 15.
Feb 10 11:02:07 host kernel: Kernel logging (proc) stopped.
Feb 10 11:02:07 host rsyslogd: [origin software="rsyslogd"
swVersion="5.6.3" x-pid="15884" x-info="http://www.rsyslog.com";]
start
Feb 10 11:02:07 host kernel: imklog 5.6.3, log source = /proc/kmsg
started.
Here is a better representation of it:
http://pastebin.com/M6Lwseqr
the two '/var/log/everything_to_send' where being used for
debugging. I
actually stacked both of those lines directly above the
@logserver.colorado.edu line, and that's when I noticed a
discrepancy
between what was getting logged to everything_to_send and
everything_to_send2. Then I bumped those lines back up above the
queuing
directives and observed that debug (and a bunch of other filtered
messages) were making it to logserver.colorado.edu.
-- Eric
Rainer Gerhards wrote:
HAve you checked for any error messages from rsyslogd inside your
logs (do
you even log them -- many distros don't do by default :-(( )
I am asking because the snippet looks syntactically wrong (then on
a
separate
line).
Rainer
-----Original Message-----
From: rsyslog-boun...@lists.adiscon.com [mailto:rsyslog-
boun...@lists.adiscon.com] On Behalf Of Eric Schoeller
Sent: Thursday, February 10, 2011 4:51 PM
To: rsyslog@lists.adiscon.com
Subject: [rsyslog] Queuing subsystem and message filtering
Hello list,
I recently tried out message queuing using rsyslog-5.6.3.
Take the following config snippet:
...
# Throw out all messages with debug severity before we log to the
network
if $syslogseverity-text == 'debug'
then ~
$WorkDirectory /var/spool/rsyslog # location for work
(spool)
files
$ActionQueueType LinkedList # use asynchronous
processing
$ActionQueueFileName logserver-queue # set file name, also
enables
disk mode
$ActionResumeRetryCount -1 # infinite retries on
insert
failure
$ActionQueueSaveOnShutdown on # save in-memory data if
rsyslog
shuts down
# Log anything that hasn't been specifically filtered out with '~'
to
logserver
*.* @logserver.colorado.edu
...
In this scenario I would assume that all messages with the
severity
of
debug don't get logged over the network to
'logserver.colorado.edu'.
But
they do. If I comment all 5 of the directives that activate
queuing,
the
debug messages are no longer logged over the network.
Is this by design, and if so could someone explain why? Is this a
bug?
Do I have a configuration problem?
Thanks!
Eric Schoeller
University of Colorado, Boulder
Information Technology Services
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com
