Skip to end of banner
Go to start of banner

Copy of opEvents - Syslog Handling - Adding a New Vendor

Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

Purpose

This page will explain how to add a new node vendor in the event the default settings are not handling the syslog traps properly. 

For this discussion well use the term 'newVendor' to be the variable that represents the new vendor we want opEvents to handle.

Steps

  • Choose a unique syslog facility for the newVendor.
  • Provision rsyslog to handle the traps appropriately.
  • Provision opEvents to parse and process the traps.

rsyslog Provisioning

Determine what facility level these syslog traps should be stamped with.  The syslog server will key on this facility level in order to send the syslog trap to the proper file.  If the device syslog is very similar to Cisco then you may want to simply use the local7 facility and the syslog traps will be sent to /usr/local/nmis8/logs/cisco.log.  Configure the nodes in question to send syslog to NMIS at the proper facility level.  For this  example we will use local6 for newVendor.  Typically facilities local0 through local7 are used for processing syslog from external nodes.

Ensure the syslog server is provisioned to received traps (udp & tcp).  This configuration is below and can be made on the /etc/rsyslog.conf file.

### /etc/rsyslog.conf

# enable network sources
module(load="imudp")
input(type="imudp" port="514")

module(load="imtcp" MaxSessions="1000" MaxListeners="50")
input(type="imtcp" port="514"

# and handle inbound/slave NMIS syslogs
local7.*                /usr/local/nmis8/logs/cisco.log
local1.*                /usr/local/nmis8/logs/slave_event.log

 

Next we'll tell rsyslog where to file messages that arrive with the facility local6.

### /etc/rsyslog.conf

# and handle inbound/slave NMIS syslogs
local7.*                /usr/local/nmis8/logs/cisco.log
local6.*                /usr/local/nmis8/logs/newVendor.log
local1.*                /usr/local/nmis8/logs/slave_event.log

 

After modifying /etc/rsyslog.conf the syslog daemon must be restarted.

[root@opmantek rsyslog.d]# /etc/init.d/rsyslog restart
Shutting down system logger:                               [  OK  ]
Starting system logger:                                    [  OK  ]

Now when syslog traps are received with facility level local6 we will see them in the /user/local/nmis8/logs/newVendor.log file.  If this file does not exist it will be created automatically.

opEvents Provisioning

For the sake of this discussion let's assume the new vendor can be parsed with the existing cisco_alternate rules found in /usr/local/omk/conf/EventParserRules.nmis.  We need to tell opEvents to use these parser rules on /usr/local/nmis8/logs/newVendor.log.  This is done by modifying /usr/local/omk/conf/opCommon.nmis.  Find the 'opevents_logs section and add the 'cisco_alternate', '<nmis_logs>/newVendor' relationship.

### /usr/local/omg/conf/opCommon.nmis

    'opevents_logs' => {
      'cisco_alternate' => [
        '<nmis_logs>/newVendor.log'
      ],
      'cisco_syslog' => [
        '<nmis_logs>/cisco.log'
      ],
      'nmis_eventlog' => [
        '<nmis_logs>/event.log'
      ],

After modifying opCommon.nmis the opEvents daemon must be restarted.

[root@opmantek ~]# /etc/init.d/opeventsd restart
Restarting opevents daemon opeventsd                       [  OK  ]
[root@opmantek ~]# 

Create an event action policy as described here: Event Actions and Escalation

Once these actions are complete the syslog traps from newVendor should be seen in opEvents.

 

  • No labels