may be problems with variable names that start with _ I seem to
remember seeing some patches around this and the imjournal module recently.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/pr
:-)
Also, with rsyslog v7+ the config optimizer makes it so that if..then tests are
just as fast as the :var, test ones. so just pick what makes the resulting
config the easiest to understand
David Lang
On Fri, 15 Apr 2016, Stuart Kendrick wrote:
Ahhh, OK.
So this config line does suppres
Just a note that the cron jobs to generate PPAs are failing due to the
cleanup changes that require a PPA variable be defined in the config
David Lang
On Mon, 18 Apr 2016, Rainer Gerhards wrote:
Hi all,
I have just placed pre-release tarballs for tomorrow's upcoming
rsyslog 8.18.0 re
it
that can cause these sorts of corruptions (this is why libfastjson was forked
from it)
David Lang
On Mon, 18 Apr 2016, Alec Swan wrote:
Hi there,
I've been seeing the following stack trace in /var/log/messages. It seems
like something is trying to shut down rsyslog causing it to leav
of the OOM handler has changed over
time, so it may vary depending on what you are running.
David Lang
On Tue, 19 Apr 2016, singh.janmejay wrote:
Won't oom be a sigkill?
You can determine which process issued term (syscall is kill).
Match it with time of death to find the culprit.
On Apr
I know that there is the ability to format a timestamp as a unix time value, but
is there a way to take a variable that contains the unix time value and output
it as a human readable string?
i.e. if $!foo = "1454527741918"
output "Feb 23 11:51:
tjson, and that fixes quite
a number of known segfaulting issues.
David Lang
On Wed, 20 Apr 2016, Christopher Racky wrote:
Date: Wed, 20 Apr 2016 14:18:25 +0200
From: Christopher Racky
Reply-To: rsyslog-users
To: rsyslog-users
Subject: [rsyslog] segfault with rsyslog Version > 8.16.
for what little it's worth, I've been running the 8.18 git tree for a couple
weeks and it's been better than the prior ones. But I am using the git version
of libfastjson
David Lang
On Wed, 20 Apr 2016, Brian Knox wrote:
Date: Wed, 20 Apr 2016 16:33:56 +
From: Brian
d approach, fix rsyslog to provide at-least-once
guarantees or fix documentation so it mentions that rsyslog doesn't provide
it?
I think it should be at least once. It's a lot easier to weed out duplicates
than to notice that something
Rainer, ping on this issue.
David Lang
On Tue, 29 Mar 2016, Kane Kim wrote:
Date: Tue, 29 Mar 2016 10:46:49 -0700
From: Kane Kim
Reply-To: rsyslog-users
To: rsyslog-users
Subject: Re: [rsyslog] retry if output module returns RS_RET_SUSPENDED
Updated config to:
module(load="omrelp&quo
but you were busy at the time and were never able to follow up on the topic.
David Lang
On Wed, 20 Apr 2016, Rainer Gerhards wrote:
Date: Wed, 20 Apr 2016 21:01:39 +0200
From: Rainer Gerhards
Reply-To: rsyslog-users
To: rsyslog-users
Subject: Re: [rsyslog] rsyslog's batch api doc
rrors.
David Lang
On Wed, 20 Apr 2016, Rainer Gerhards wrote:
I should also say that it had a release goal, so it is on my active to-do
list.
Sent from phone, thus brief.
Am 20.04.2016 19:40 schrieb "Rainer Gerhards" :
Yup... Look aty work Log - still no time... Was meant for last r
upt in
it, not an entire batch.
David Lang
On Wed, 20 Apr 2016, Kane Kim wrote:
What would happen if it will be retried forever at this point? Would it
affect any other queues or only current action queue. If it will not affect
anything else I'd really prefer it to block there until successfu
ld be satisfied with a simple retry/reconnect, but I think we should get
things correct first, and then look to optimize it.
David Lang
On Wed, 20 Apr 2016, David Lang wrote:
Date: Wed, 20 Apr 2016 23:44:51 -0700 (PDT)
From: David Lang
Reply-To: rsyslog-users
To: rsyslog-users
Subject: Re: [rsys
legit for the
module to just say "temp problem" and tear down the connection and leave the
retry up to the core logic.
David Lang
On Wed, Apr 20, 2016 at 11:44 PM, David Lang wrote:
The original discussion when we created the batch mode was that there were
two possible ways for
depending on how you look at it, it's serialized or
parallel :-)
in v7, the worker thread processed all messages for each action before moving
to the next action. But that caused problems with things like global variables
where you want to process all actions with one value (and one log mes
and 2, only for step 0 (start batch) and
step 3 (mark batch as done either succeeded/failed)
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? F
x27;t work (global
variables, counter modules, etc)
David Lang
On Thu, 21 Apr 2016, David Lang wrote:
On Thu, 21 Apr 2016, Kane Kim wrote:
Thanks for great explanation, David, that really helped me to understand
this part. What I'm suggesting is essentially this (and please correct me
;ll
write up a more formal conceptual document with diagrams so that we can have
this all documented for the next time we have to look at it in the future :-)
David Lang
That's also one of the main things for
me to take time (memory doesn't server well if you need to be very
specifi
s from the adiscon repo (no
stray packages with the same name/version from redhat repos)
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslo
On Thu, 21 Apr 2016, Rainer Gerhards wrote:
2016-04-21 9:57 GMT+02:00 David Lang :
On Thu, 21 Apr 2016, Rainer Gerhards wrote:
2016-04-21 9:35 GMT+02:00 Kane Kim :
Thanks for great explanation, David, that really helped me to understand
this part. What I'm suggesting is essentially
On Thu, 21 Apr 2016, Rainer Gerhards wrote:
2016-04-21 10:13 GMT+02:00 David Lang :
On Thu, 21 Apr 2016, Christopher Racky wrote:
Hi,
Thanks a lot so far.
David you assumed right. I was not able to get a coredump as only one
thread dies.
I noticed, that on my RHEL 6.7 System I have json-c
rsyslog should handle it just fine.
Rsyslog detects that the inode of the filehas changed and starts from the
beginning of the new file.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com
written to it before rsyslog reads the file again, rsyslog
could miss the fact that the file was truncated and miss data and/or spit out a
partial log.
David Lang
On Fri, 22 Apr 2016, Ashish Barmase wrote:
Date: Fri, 22 Apr 2016 17:08:18 +
From: Ashish Barmase
Reply-To: rsyslog-users
t accepted, consider removing or
replacing it
are you mixing the legacy and new style syntax? can you show us your config?
David Lang
On Mon, Apr 25, 2016 at 12:00 PM, Muhammad Asif
wrote:
Dear Ashish, Thanks for reply. How you did this exactally. Any cron job or
some thing else?
Regards
M.As
to
/var/spool/rsyslog? especially as user user/group syslog.syslog?
as far as rotating the file go, you should not copy+delete the file, you should
move the file, then HUP rsyslog and rsyslog will recreate the file as needed.
David Lang
On Mon, 25 Apr 2016, Muhammad Asif wrote:
Date: Mon
rsyslog 5.8 is ancient (8.18 is current), there should be a 7.x version
available through the redhat repos
I have never used the aureport tool, but from the example on the page, you can
give it the -f flag to specify a filename to look at.
David Lang
On Thu, 28 Apr 2016, Warron S French
are you sure the Cisco is supporting TLS, not just doing plaintext over TCP?
If the client has a cert to offer, we still want to gather the data from it,
even if we don't require validation.
David Lang
On Mon, 2 May 2016, Wingard, Nathaniel wrote:
Date: Mon, 2 May 2016 17:35:20 +
of to make sure that
rsyslog has the correct socket open after the HUP
If this doesn't work, a full restart will be needed.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-service
arrive via tcp or udp using the rules in the
inbound ruleset (and using a separate queue)
David Lang
On Wed, 4 May 2016 16:38:34 +, Singh, Radesh wrote:
Hello,
Perhaps I'm overthinking this, but as I've confused myself pretty
good, I'm reaching out to you guys.
We noticed that
you show your full config?
David Lang
On Thu, 5 May 2016, Kuo Hugo wrote:
Date: Thu, 5 May 2016 01:28:38 +0800
From: Kuo Hugo
Reply-To: rsyslog-users
To: rsyslog@lists.adiscon.com
Subject: [rsyslog] [Log delayed] Some old logs appears in the log file
Hi Folks,
I observed a weird behaviour
you need to use one syntax or the other
$UDPServerRun 514
is the same as
input(type="imudp" port="514")
so remove the ServerRun lines and I think you have it. (although I think you
should put the two input() statements on different lines)
David Lang
On Wed, 4 May 2016,
Yes, you need to be using at least 7.x (current is 8.18)
David Lang
On Wed, 4 May 2016, Singh, Radesh wrote:
Date: Wed, 4 May 2016 19:32:18 +
From: "Singh, Radesh"
Reply-To: rsyslog-users
To: rsyslog-users
Subject: Re: [rsyslog] Remote messages getting into local logs
Rsysl
y run into problems where the
answers that we give are not going to work.
David Lang
On Wed, 4 May 2016, Singh, Radesh wrote:
Date: Wed, 4 May 2016 19:42:46 +
From: "Singh, Radesh"
Reply-To: rsyslog-users
To: rsyslog-users
Subject: Re: [rsyslog] Remote messages getting into local lo
the message that you have received and then create a custom
template to output the message with the extracted value in the msgid field.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com
Even though the RFC allows it, it's hard to imagine a valid log without a
timestamp or source.
David Lang
On Fri, 6 May 2016, Rainer Gerhards wrote:
The best way to diagnose these kinds of problems is to add
*.* /var/log/msgdebug;RSYSLOG_DebugFormat
To the top of rsyslog.conf. The
logs.
David Lang
On Fri, 6 May 2016, Zhou, Tao wrote:
Thanks David for youy prompt reply.
Did you mean rfc5424 ? if so, the message is deifned in BNF:
SYSLOG-MSG = HEADER SP STRUCTURED-DATA [SP MSG]
HEADER = PRI VERSION SP TIMESTAMP SP HOSTNAME SP APP-NAME SP PROCID SP
MSGI
in-memory queue)?
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBLIC mailing list
tead of 115-186-137-215.nayatel.pk.
that depends on what data you have and where it came from.
The default template uses $hostname in that field, but if the sender is
violating the RFCs by not providing a hostname, rsyslog will make it's best
guess by doing a reverse lookup ($fromhost) of th
I thought the last time the subject came up he was saying he needed help on
this.
I was actually going to ping you today to see if you had been able to get any
time to look at it :-)
David Lang
On Wed, 11 May 2016, Kane Kim wrote:
Date: Wed, 11 May 2016 10:46:52 -0700
From: Kane Kim
queue.dequeuebatchsize = "1000" queue.size="50" queue.filename="x_dc_queue"
action.resumeRetryCount="-1" queue.maxfilesize="256m" queue.saveonshutdown="on")
David Lang
Rainer
2016-05-11 16:21 GMT+02:00 David Lang :
I had a short o
ring module for json output.
2. sanitize mm module so that I can stick data into ES.
David Lang
On Wed, 11 May 2016, Kane Kim wrote:
Sorry guys, I was busy as well for a couple of weeks, let's reiterate what
we will have to do here. I guess as Rainer said first step should be to
write a
t it can make it so that you can't read the
file.
Just to be sure I understand your situation, normal logs from both files work,
it's the imfile section that works on one and not on the other, correct?
you can try starting rsyslog in debug mode (rsyslogd -dn) and see if the log
giv
to find the answers.
David Lang
On Thu, May 12, 2016 at 2:12 PM, David Lang wrote:
On Thu, 12 May 2016, Thomas Lowry wrote:
Hello,
I am having an issue where I have two identical servers and a central log
server. We will call the identical servers "server1" and "server2"
dFormat
$ActionExecOnlyWhenPreviousIsSuspended off
With the new format, queues and rulesets are really easy
ruleset(name="foo" queue.type="linkedlist" ...) {
*.info :omrelp:syslog1.example.com:1514;RSYSLOG_ForwardFormat
$ActionExecOnlyWhenPreviousIsSuspen
That general approach is what I recommend in any case.
David Lang
On Thu, 12 May 2016, Joe Blow wrote:
This probably won't answer your exact question but could you just use
keepalived and create a VIP to send the syslog/relp to? Would there be any
downside to using a vrrp pool for
ble to you
through an update
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBL
yslogd -N3" or similar
2. log with the format RSYSLOG_DebugFormat so that you can see what variables
contain what values so that you can check that your filters match the variable
contents.
David Lang
On Mon, 16 May 2016, J. Toby Knudsen wrote:
Date: Mon, 16 May 2016 12:10:00 -0400
From:
not most of the time. occasionally when I'm testing something (although not for
the last few months as I've needed liblognorm2 that wasn't being shipped yet)
David Lang
On Mon, 23 May 2016, Rainer Gerhards wrote:
Hi folks,
I wonder if someone here from the list uses the da
clude everything since at least one month).
is it the libfastjson that's the question or the liblognorm v2?
David Lang
Hope that clarifies and thanks for the comments. I think I'll postpone
if nobody says he would really like to have it with next weeks release
(holding rsyslog for th
he options that we really should support and provide some sort of
explination as to what they mean pretty easily.
Unfortunatly this is why so many TLS related questions go unanswered for a while
here on the list.
David Lang
___
rsyslog ma
e lisa presentation and the first article in my logging
series:
https://www.usenix.org/conference/lisa12/technical-sessions/presentation/lang_david
https://www.usenix.org/publications/login/david-lang-series
https://www.usenix.org/publications/login/april14/lang
David Lang
debug log locally (not necessarily the full messages, timestamp
would be enough)
you can then see if the local file in growing while things are not yet showing
up in ES to see if the issue is on the sending side or on the receiving side.
David Lang
On Tue, 24 May 2016, Alec Swan wrote:
Date
say, audits usually don't care about real risk, they have their
checkboxes...
David Lang
On Tue, 24 May 2016, David Lang wrote:
Date: Tue, 24 May 2016 16:16:22 -0700 (PDT)
From: David Lang
To: rsyslog-users
Subject: Re: [rsyslog] Setting TLS cipher suite / disabling RC4
On Tue, 24 May 201
On Wed, 25 May 2016, Thomas Güttler wrote:
Am 25.05.2016 um 08:26 schrieb Rainer Gerhards:
2016-05-25 8:22 GMT+02:00 Thomas Güttler :
Am 24.05.2016 um 17:40 schrieb Rainer Gerhards:
2016-05-24 17:10 GMT+02:00 Thomas Güttler :
Hi,
I have some remote hosts which can't connect to our central
-site VPN.
In either case, you will want to use RELP instead of just TCP
http://blog.gerhards.net/2008/04/on-unreliability-of-plain-tcp-syslog.html
http://blog.gerhards.net/2008/05/why-you-cant-build-reliable-tcp.html
David Lang
I will use RELP, thank you for this hint.
Since it is based on tcp
On Thu, 26 May 2016, Karri Vrkreddy wrote:
Thanks David Lang for looking into this.
Please correct me if I understood your suggestion incorrectly.
I have changed earlier configuration to
***
module(load="imudp")
ruleset(name="central"
queue.type="Link
m on it's stdin.
What your program does after that is strictly up to it.
If you are wanting to just modify the message, you shouldn't use omprog, you
should use mmexternal which lets you modify the log message in rsyslog with an
external program and then
ur perl program to write it's stdin to a file and
make sure that what you get from rsyslog is what you think you are getting.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-s
Anyone have any suggestions (even quick-n-dirty ones) to let this work?
David Lang
/usr/src/libfastjson# make dist
make dist-gzip am__post_remove_distdir='@:'
make[1]: Entering directory `/usr/src/libfastjson'
if test -d "libfastjson-0.99.3.master"; then find "lib
On Wed, 1 Jun 2016, Thomas D. wrote:
OK, workaround:
Run "make" before "make dist".
I had to do make clean; ./configure; make; make dist
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
ht
the output for
impstats to see how many messages are queued.
David Lang
Thanks,
Alec
On Sat, May 28, 2016 at 2:40 AM, Rainer Gerhards
wrote:
2016-05-28 4:36 GMT+02:00 Alec Swan :
I ran rsyslog in debug mode and now I am in the same state where a copy
of
the logs are still being writt
og started back up with 925MB RAM usage. What
parameter controls how much of queue data rsyslog will load up from disk on
startup?
It reads old messages in it as it has time, there is no specific parameter. I
don't know what the chunk size is that it processes.
David Lang
Thanks,
Alec
set. You can change this by editing the source, but above some point you will
trip over other issues (assumptions about how large a message can be influencing
buffer sizes, etc)
you can try it, but *syslog is designed for messages of a few KB, so you are
~1000x the design size.
David Lang
, but it's not the right tool
to use with a screw. You may be able to get it to work, but it's still not the
right tool.
rsync or similar is the right answer when you have many files in the MB range to
get to another system, not *syslo
using the new syntax, action.resumeRetryCount="-1" works on a queue for an
action, but doesn't work for a queue on a ruleset, resumeRetryCount="-1" also
doesn't work.
What should work?
David Lang
___
rsyslog mailing
would make sense to
try and globally define parsers in that case.
David Lang
regards
Chris
2016-06-16 15:55 GMT+02:00 Rainer Gerhards :
You need to specify them in the ruleset object, e.g.
ruleset(name="remote" parser=["rsyslog.srcipinject","rsyslog.lastline&
and then notice that the inode for the filename has changed (i.e. that
the move happened), and close the old file and start reading the new one.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.c
it doesn't require changing the source
code of the logging library.
Ok, just be aware that you will loose some logs as they rotate.
what version of rsyslog are you running? current versions should not require a
HUP.
David Lang
Thanks,
Alec
On Thu, Jun 16, 2016 at 11:36 AM, David Lang
looking at:
http://www.rsyslog.com/doc/v8-stable/configuration/modules/imfile.html
did you try the reopenOnTruncate option?
David Lang
On Thu, 16 Jun 2016, Alec Swan wrote:
I am using rsyslog version 8.19. It did require restart (I haven't tried
HUP yet) after logrotate, but I wond
https://www.usenix.org/publications/login/david-lang-series
https://www.usenix.org/conference/lisa12/technical-sessions/presentation/lang_david
you can use the (non-free) rsyslog windows agent, or nxlog to gather logs from
windows and deliver them to central rsyslog servers. look through the
nt to have 50 different ways to handle the same Cisco message, but
how do we pick which of the many different versions we are going to use?
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/pr
21
7e 154MB 25
8e 152MB 35
9e 150MB 31
David Lang
I'd love to see the audispatch work you've done.
Also, in case you were wondering, I have a patent on my log-to-jpg-to-sql
module, and the corresponding OCR extraction tool.
Cheers,
JB
Original Message
From:warron.fre...@
hings haven't changed)
At this point I've basically abandoned bzip2, I don't fine any cases where it's
less cpu or more compression than is available via xz.
About the only time I still use gzip is for the logs compressed by rsyslog.
David Lang
When testing with firewall
logs
On Wed, 22 Jun 2016, warron.french wrote:
David Lang, you also have valid points obviously, but what I wonder is if
the server crashes where is the data going to be then. At least one
centralized on a NAS/SAN solution a massive hit to the server won't be as
destruction to the data on th
On Wed, 22 Jun 2016, Joe Blow wrote:
If you're trying for true HA/prod setup, I'd suggest looking at
DRBD+pacemaker+corosync+crmsh+keepalived+rsyslog with local storage.
Why would use keepalived as well as pacemaker/corosync? It seems to me that you
would use one or the other.
I've given s
me. By then having
one copied to the other based on which is the 'primary' I've got a rsyncable
master 99%+ of the time, and the remainder of the time I have the individual
files that I can inspect and merge manually if it absolutly matters (it seldom
does so I just pick one a
build the list of culprits? or the best way to
deal with them once you have the list?
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? F
e the
list.
given the number of times this has come up, I'm sure there is some interest.
Thanks for volunteering on this.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
the comfort to start using
the parsing.
Given that liblognorm is pretty insensitive to the number of rules, it may be
that we can craft a combined rulebase that could ship by default with liblognorm
as a starter for people.
David Lang
___
rsyslog mai
ntain.
David Lang
Philippe
-Original Message-
From: rsyslog-boun...@lists.adiscon.com [mailto:rsyslog-
boun...@lists.adiscon.com] On Behalf Of David Lang
Sent: Thursday, June 23, 2016 7:08 PM
To: rsyslog-users
Subject: Re: [rsyslog] Dealing with malformed messages
On Thu, 23 Jun
the feature, it can be available pretty quickly.
David Lang
On Tue, 28 Jun 2016, Alec Swan wrote:
Hello,
I am using rsyslog 8.19.0 with mmnorlamize and liblognorn to parse our log
messages and convert them to JSON. After that I use omelaticsearch to send
JSON to Elasticsearch 2.2.3.
I noti
ding sure seems like it's
complaining about '.' (which is ascii 46)
do you have any way of finding out what the full string is that it's trying to
process?
have you asked elastic.io? (since the problem seems to be in their end), even if
it's invalid json, they shou
file that has the Hindi characters near the beginning of it and try the file -bi
again, see if it gives a more accurate answer.
otherwise, you will have to track down what's writing the messages and try to
set the character set there (or at least find out what character set it's using)
This is helping narrow things down.
I would have rsyslog write to a file with the template that you use to send to
elasticsearch.
I would also use the liblognorm command-line tool to parse the file and output
json.
let's try to see where it breaks.
David Lang
On Wed, 29 Jun 2016,
other than that,
it will continue to work.
the 'new way' of doing this would be:
if ($programname == "named") and ($pri-text == "daemon.info") then {
/var/log/syslog/named
stop
}
use whichever way makes the most sense to you
David Lang
_
On Thu, 30 Jun 2016, Marc Haber wrote:
On Thu, Jun 30, 2016 at 12:59:25AM -0700, David Lang wrote:
if ($programname == "named") and ($pri-text == "daemon.info") then {
/var/log/syslog/named
stop
}
Thanks, that was helpful.
The docs need more examples.
patches we
are you getting any logs into rsyslog?
do the mail.* logs show up in the systemd journal and just not in rsyslog?
The first thing to do is to track down where things fail before starting to dig
into the config files of any one thing.
David Lang
On Thu, 30 Jun 2016, Adrian P. van Bloois
On Fri, 1 Jul 2016, Marc Haber wrote:
On Thu, Jun 30, 2016 at 10:46:27AM -0700, David Lang wrote:
On Thu, 30 Jun 2016, Marc Haber wrote:
On Thu, Jun 30, 2016 at 12:59:25AM -0700, David Lang wrote:
if ($programname == "named") and ($pri-text == "daemon.info") then {
/
Linux logs for such failures.
David Lang
On Sun, 3 Jul 2016, Dr. Yury Zhauniarovich wrote:
Date: Sun, 3 Jul 2016 16:48:49 +
From: Dr. Yury Zhauniarovich
Reply-To: rsyslog-users
To: "rsyslog@lists.adiscon.com"
Subject: [rsyslog] Strange behavior of rsyslog, once it is run as
first I've heard of this, we'll need to have Rainer comment on this.
David Lang
On Wed, 6 Jul 2016, Alec Swan wrote:
Date: Wed, 6 Jul 2016 15:34:44 -0600
From: Alec Swan
Reply-To: rsyslog-users
To: rsyslog-users
Subject: Re: [rsyslog] Invalid JSON from
mmnormalize/
themselves are corrupt.
I am not aware of the exact details of the 8.18 improvements
try the recover_qi.pl script (it's been posted before, I'm attempting to include
it as an attachment here), see what it has to say about the queue files.
David Lang#!/usr/bin/perl -w
# recover rsyslog
As I understand Rainer's reply, if you compile from the current liblognorm
master you will not have that limitation. This new branch will be released
shortly (along with rsyslog 8.20)
David Lang
On Thu, 7 Jul 2016, Alec Swan wrote:
Date: Thu, 7 Jul 2016 12:57:51 -0600
From: Alec Swan
Per Rainer, the 10k limit is in the normalizer tool, not in the library. In
rsyslog there is a maxmessagesize parameter. That is the limiting factor even
with the existing version.
the new version of liblognorm/mmnormalize will work with existing version 1
rulesets.
David Lang
On Thu, 7
col = "ipv4-only"
net.enableDNS = "off"
parser.escapeControlCharactersOnReceive = "on"
)
I don't know how ES could be complaining about a 50K message if it's defaulting
to 8K
David Lang
On Thu, 7 Jul 2016, Alec Swan wrote:
I remember seeing MaxMessageSize parameter, but
(at which point rsyslog may
try to flush the memory queues to disk if that's how you have it configured),
and then if rsyslog doesn't shutdown fast enough (5 seconds or so), issues a
kill -9. If there is too much data to flush to disk, this can cause the .qi file
to not
ot found online. The full config is at
http://pastebin.com/ENkzmMT2
What have I done wrongly?
what version of rsyslog are you running?
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/pr
.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED
Also note as Rainer posted in a separate thread, binary packages of 8.20 are
going to be delayed while work is done to figure out how to avoid the package
problems of the last few days for future releases.
David Lang
On Tue, 12 Jul 2016, Florian Riedl wrote:
Date: Tue, 12 Jul 2016 18:09:46
s got fixed in < 10 hours, but also
surprised if they didn't fix it within 18 hours or so, just due to time zone
effects.
David Lang
___
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/profess
501 - 600 of 3956 matches
Mail list logo