Ad Widget
Collapse
Better syslog message handling for Zabbix
Collapse
X
-
-
zbxlog release 0.6
Hi,
I have just released zbxlog v0.6 (http://www.alixen.org/attachments/do...bxlog-r0.6.tgz).
This release includes some enhancements:
- frontend patches for Zabbix 1.8.4
- Don't exit on temporary DB access error (new parameter: db_error_max_retries)
- Use remote IP as remote hostname when hostname undefined (no reverse DNS)
- Replace 'Zabbix agent (active)' by 'Zabbix Trapper' item type (conflict with real agent)
- added chmod +x /etc/init.d/zbxlog in installation instructions
Happy syslog monitoring !
AlixenComment
-
Thank you Alixen for this new release
I finally found where was the problem by using debug mode (set "my $DEBUG = 1" in all files from zbxlog/lib/Zbxlog)
The server wasn't able to do reverse DNS resolution on its own ip address.
So I added to /etc/hosts :
192.168.10.27 SVIMOCAVL007Comment
-
Hi,
Not by me.
We don't have any Oracle server in our environment.
However, Zbxlog uses Perl DBI to access Zabbix database.
If you can install Perl DBD::Oracle package, configure zbxlog to use it (set in etc/zbxlog.conf: db_type=Oracle) and test it, please report any error message here. I'll try to help you.
Regards,
AlixenComment
-
zbxlog release 0.7
Hi,
I have just released zbxlog v0.7 (http://www.alixen.org/attachments/do...bxlog-r0.7.tgz).
This release :- adds preliminary support for upcoming Zabbix 2.0 (tested on 1.9.1)
- has been tested on Ubuntu Linux 10.04 LTS
It works with zabbix 1.9.1 (and of course with 1.8).
PHP frontend patches are not available yet for 1.9.1, message severity appears as "Unknown".
Happy syslog monitoring !
JBoComment
-
zbxlog release 0.8
Hi,
I have just released zbxlog v0.8 (http://www.alixen.org/attachments/do...bxlog-r0.8.tgz).
Sorry to release it so soon after r0.7 but 1.9.1 frontend patches were much easier than I thought.
This release adds:- Zabbix 1.9.1 frontend patches
- No exit on temporary zabbix server connection error (new parameter: zabbix_error_max_retries)
- Better database error reporting messages
This release :- is feature complete : syslog message handling has not changed since 0.5.
- is resilient to temporary zabbix server or database connection errors.
- supports current Zabbix production version (1.8) and future 2.0 release (1.9 branch).
All users of Zbxlog are urged to upgrade to this release and test it.
Happy syslog monitoring !
JBoComment
-
Hi all,
Im trying to install zbxlog 0.8 in zabbix 1.8.4 using postgresql. I think the zbxlog part, I managed to correctly configure it and run since from the log I get "zbxlog.pl" started and verified via ps auxwww|grep zbxlog.
But the zabbix configuration part I am not sure if I did it correctly.
Is there a visual step by step guide with regards to creating a sample zbxlog Item, Trigger Definition, and the sample result. Will it also be possible to say graph the gathered data? I am also not sure how to check that syslogs are getting instered and processed in the zabbix database?
Here is what I did,
On the host "A Test Host", I created a sample Item (zbxlog_item.png) with the following information:
Description: Syslog
key: syslog[]
I also created a trigger (zbxlog_trigger.jpg) with the following information
name: Syslog Debug
expression: {A Test Host:syslog[].logseverity(0)}>10
I am not sure if what I did is correct, since I am not getting any Syslog information from the "latest data" menu (zbxlog_latestdata.jpg) , nothing in the Last Check, Last Value and Change Columns.
TIA
Comment
-
Hi,
You need to make your item work before trying to set a trigger.
zbxlog needs to find Zabbix host name (here "A Test Host") based on its IP address.
Does "A Test Host" have a correct IP address ?
zbxlog silently drops all syslog messages that have no corresponding host/item set up unless you create a specific "no match" host.
"no match" host and item are defined in zbxlog.conf:
I suggest that you create an host named sysloghost with a single Zabbix trapper item named syslotg_nomatch with type Log.Code:# syslog messages not matched by any syslog[] item will be sent to # nomatch_item on nomatch_host # messages will be discarded if nomatch_host is empty nomatch_host=sysloghost nomatch_item=syslog_nomatch
Actual content of this item may give a clue on why your setup doesn't work.
Hope this helps,
JBoComment
-
Hi,
Zbxlog with Postgres is now working.
No need for a new version, just make sure DBD::Pg is installed and zbxlog DB connection is correctly configured.
I will add this information in a future release.
In the meantime here what I have done (on a Debian 6.0 Squeeze):
Install perl DBI and DBD:Pg modules:
Define Postgres connection in /usr/local/zbxlog/etc/zbxlog.confCode:apt-get install libdbi-perl apt-get install libdbd-pg-perl
Make sure that Zbxlog is using Zabbix 1.8 DB schema:Code:# Zabbix database connection parameters db_type=Pg db_name=zabbix db_host=localhost db_port=5432 db_user=zabbix db_password=******* (get it from /etc/zabbix/zabbix_server.conf : DBpassword)
Check /usr/local/zbxlog/var/zbxlog.log for error messages.Code:# Zabbix server version # Supported values: # zabbix_version=1.8 # zabbix_version=2.0 zabbix_version=1.8
Hope this helps,
JBoComment
-
Zbxlog init script for Centos
Hello,
I just created an init script for Centos:
I hope it would be useful for someone else
/etc/init.d/zbxlog
Code:#!/bin/sh # This is the init script for starting up the # Zbxlog daemon # # description: Starts and stops the Zbxlog daemon. # processname: zbxlog.pl # # Source function library. . /etc/rc.d/init.d/functions # Get config. . /etc/sysconfig/network # Check that networking is up. [ "${NETWORKING}" = "no" ] && exit 0 Zbxlogdir=/usr/local/zbxlog procname="zbxlog.pl" start(){ echo -n "Starting $procname service:" daemon "$Zbxlogdir/bin/zbxlog.pl &" RETVAL=$? echo [ $RETVAL -eq 0 ] && touch /var/lock/subsys/$procname return $RETVAL } stop(){ echo -n "Stopping $procname service:" killproc $procname RETVAL=$? echo [ $RETVAL -eq 0 ] && rm -f /var/lock/subsys/$procname return $RETVAL } restart(){ stop start } status(){ numproc=`ps -ef | grep 'perl' | grep $procname | grep -v grep | wc -l` if [ "$numproc" -gt 0 ]; then echo "$procname is running" else echo "$procname is stopped" fi } # See how we were called. case "$1" in start) start ;; stop) stop ;; status) status ;; restart) restart ;; *) echo "Usage: $0 {start|stop|status|restart}" exit 1 esac exit $RETVALComment
-
Comment