Configuring Infrastructure Data Collector for Clustering

Edit online

About this task

Infrastructure Data Collector supports different clustering modes. Based on your requirements, you can choose between:

  • Clustering for high availability
  • Clustering for load sharing

Follow the instructions below for setting up clustering for Infrastructure Data Collector.

Infrastructure Data Collector Clustering for High Availability

Edit online

About this task

Infrastructure Data Collector supports active/passive clustering for high availability scenarios. In this case, one of the Infrastructure Data Collector instances is marked as active, and the other one as passive. The passive Infrastructure Data Collector node is stopped, and should be started when the active node goes down.

Note: Infrastructure Data Collector does not support automatic switching from active to passive cluster node. This must be done by the user.

To configure Infrastructure Data Collector clustering for high availability

Procedure

  1. Install two Infrastructure Data Collector instances.
  2. Start the two instances.
  3. In My webMethods, configure the two Infrastructure Data Collector instances.
    Important: Make sure you register the two instances with the same collector name under Collector Settings for Infrastructure Data Collector.

    For more information about defining collector settings for Infrastructure Data Collector, see Configuring BAM.

  4. Shut down the two Infrastructure Data Collector instances.
  5. Start the active Infrastructure Data Collector instance. It discovers assets and monitors components, as configured.
  6. To test the clustering scenario, shut down the active Infrastructure Data Collector node, and start the passive one. The monitoring should continue as usual.
    Note: If you are monitoring SNMP agents, you have to re-discover all of them on the passive Infrastructure Data Collector node.

Infrastructure Data Collector Clustering for Load Sharing

Edit online

About this task

Infrastructure Data Collector supports clustering for load sharing. This is useful in case your system contains a large number of assets and components, or you wish to position Infrastructure Data Collector closer to the components that you are monitoring. With this clustering mode it is possible to install multiple instances of Infrastructure Data Collector and use them to collect data and send it to Analytic Engine.

To configure Infrastructure Data Collector clustering for load sharing

Procedure

  1. Install several Infrastructure Data Collector instances in your environment.
  2. Start all installed Infrastructure Data Collector instances.
  3. In My webMethods, configure the Infrastructure Data Collector instances.
    Important: On the Define Environments > Configure Servers page, under Collector Settings for Infrastructure Data Collector, make sure you register each Infrastructure Data Collector instance with a unique Collector Name, and that you point all instances to the same JMS URL.

    For more information about defining collector settings for Infrastructure Data Collector, see Configuring BAM.

  4. In My webMethods, on the Discovery Assets page, select the Infrastructure Data Collector instance that will monitor each asset.
    Note: To avoid data duplication, make sure you do not discover the same asset twice with different Infrastructure Data Collector instances.