IBM Tivoli Monitoring > Version 6.3 Fix Pack 2 > Administrator's Guide > Agent-based services > SNMP alerts > OMNIbus configuration for SNMP
IBM Tivoli Monitoring, Version 6.3 Fix Pack 2
Enable OMNIbus heartbeat automation
For Tivoli Enterprise Monitoring Agents that send situation events as SNMP alerts or EIF events to Netcool/OMNIbus, you can enable OMNIbus automation to have an event sent when the EE_HEARTBEAT is overdue. EE_HEARTBEAT life cycle status events are sent to the receiver at regular intervals to confirm that the monitoring agent is running and alerts can reach their destination.
HEARTBEAT events from SNMP and EIF are displayed on the OMNIbus event console as they arrive and the count is incremented as new events arrive from the agent.
The itm_heartbeat.sql file contains sample automations for processing the Autonomous Agent Heartbeats for both EIF and SNMP. Run the SQL file to enable the automation.
Procedure
- Copy itm_heartbeat.sql from the Tivoli Monitoring Agent DVD mibs/sample_rules/omnibus directory.
- Place the copy in the Netcool/OMNIbus installation path and run the following command:
- Where "user" is a valid user name, "password" is the corresponding password, and "server" is the ObjectServer name
%NCHOME%\bin\redist\isql.exe -U "user" -P "password" -S "server" < itm_heartbeat.sql
- Where "servername" is the ObjectServer name, "username" is a valid user name, and "psswrd01" is the corresponding password
$NCHOME/omnibus/bin/nco_sql -server servername -user username -password psswrd01 < itm_heartbeat.sql
Results
After the OMNIbus automation is installed, the automation registers the heartbeats from managed systems as they arrive. Individual heartbeats are no longer displayed and counted in the event console but, if an expected heartbeat is overdue, the automation raises a "Heartbeat missing" alert with:Summary = 'Heartbeat Missed for:' + heartbeat_missed.Node + ' last received at ' + to_char(heartbeat_missed.LastOccurrence)
What to do next
The default interval for sending the EE_HEARTBEAT status 15 minutes. You can adjust the value by modifying the interval attribute for the heartbeat status event in the trapcnfg.xml file for SNMP alerts and in the eifdest.xml file for EIF events configuration file.Especially with SNMP, one missed heartbeat does not necessarily indicate a problem, thus the default is to raise an alert after the heartbeat is overdue: (2 x heartbeat interval ) + 2 minutes. You can edit this in the itm_heartbeat.sql with the entry,
-- 2 heartbeats plus 2 minutes grace before agent missed set time_of_expiry = (new.ExpireTime * 2 * 60 + 120) + getdate();For example, add two more minutes and the setting looks like this:
-- 2 heartbeats plus 4 minutes grace before agent missed set time_of_expiry = (new.ExpireTime * 2 * 60 + 240) + getdate();
Parent topic:
OMNIbus configuration for SNMPRelated reference:
EIF life cycle event
EIF heartbeat event