Configuration setup for linear trending

To configure and monitor predictive analytics for linear trending, you require Tivoli Netcool/OMNIbus, the Probe for Tivoli EIF, and IBM Tivoli Monitoring to be installed within an integrated environment.

The following figure shows the required configuration setup for the product components in the integrated environment.

Figure 1. Tivoli Netcool/OMNIbus and IBM Tivoli Monitoring configuration for predictive events
Tivoli Netcool/OMNIbus and IBM Tivoli Monitoring configuration for linear trending

The configuration flow is as follows:

 1 
The probes are installed on the devices or systems that you want to monitor, and send the events to the ObjectServer. The automations in the ObjectServer detect the event rates for each monitored device.
 2 
The ObjectServer writes the event rate data to the event log files, and the IBM Tivoli Monitoring Health Performance Agent reads these files. To transform the events into IBM Tivoli Monitoring situations, the SQL and automations for predictive analytics must be run on the ObjectServer, as well as the appropriate situation classes. The Health Performance Analyzer feeds the situations to the Tivoli Enterprise Monitoring Server.
 3 
Tivoli Data Warehouse archives the historical event rate data. You must configure IBM Tivoli Monitoring to retain data samples in history files and save these files to the Tivoli Data Warehouse on a regular basis. Two specialized agents (the Archiving agent, and the Summarization and Pruning agent) interact with Tivoli Data Warehouse to receive, aggregate, and prune data.
 4 
Tivoli Data Warehouse feeds the archived event rate data to the Tivoli Performance Analyzer, where a trend uses the data to calculate the likely event rate in the future. Thresholds can be configured, which, when violated, generate a situation.

Tivoli Performance Analyzer consists of a configuration tool, and predefined tasks, situations, and workspaces, which are all accessible from the Tivoli Enterprise Portal. Additionally, a Performance Analyzer warehouse agent interacts with:

  • Tivoli Data Warehouse to retrieve the stored historical data collected by other agents
  • The portal server to receive the instruction to run the analytical task, and to perform analytical calculations on the data
  • The hub monitoring server to pass on the results of the trending
 5 
All situations are parsed to Tivoli Enterprise Portal. The trend can be viewed in two default workspaces.
 6 
The Tivoli Enterprise Monitoring Server forwards the situations created by Tivoli Performance Agent to the Probe for Tivoli EIF. The probe receives the situations, processes the situation attribute data, maps the data to ObjectServer fields, and then sends alerts to the ObjectServer. The probe uses the rules file configuration to convert the situation data to event data.
 7 
Events that are inserted into the alerts.status table can be viewed, filtered, and sorted in the Active Event List within the Web GUI, or in the event list. Launch-in-context functionality is also enabled from predictive events in the Active Event List, to Tivoli Enterprise Portal. This feature enables you to view details about an event in the relevant Tivoli Enterprise Portal workspace. In Tivoli Enterprise Portal, you can identify the trend to which the predictive event corresponds.