Skip to content
CA Unified Infrastructure Management - 8.5.1
Documentation powered by DocOps

Configure Discovery Queues

Last update June 22, 2017

If all discovery process probes are deployed on a single hub, communication of discovery data is automatically configured. However, if discovery probes are deployed to hubs other than the hub that hosts the discovery_server probe, ensure that discovery data can flow from those hubs up to the primary hub.

Flow is accomplished by setting up queues that handle the probe_discovery subject:

  • You can use a combination of attach and get queues.  An attach queue creates a permanent queue on a downstream hub.  A corresponding get queue set up on the upstream hub is paired with each attach queue to retrieve messages from the downstream hub.
  • Alternatively, you can use post queues.  A post queue set up on a downstream hub sends a directed stream of messages to the upstream hub.

 An attach queue is automatically set up on the primary hub to collect discovery data.  Set up additional queues to collect discovery data from downstream hubs that host discovery_agent or any CTD-publishing probes. This list includes (but is not limited to):

  • discovery_agent
  • vmware 5.10 or later
  • cm_data_import (typically deployed with discovery_server on the primary hub)
  • snmpcollector
  • vcloud
  • rhev (Red Hat Virtualization)

Then set up a corresponding get queue (which retrieves messages from the attach queue) on the primary hub and on any hub that transfers the messages to another hub.

You can set up discovery queues in either Admin Console or Infrastructure Manager.

Follow these steps:

  1. Identify the hub on which you want to create a queue and open the hub configuration GUI:
    • Admin Console: Expand the hub in the navigation tree and select its robot. Click the arrow next to the hub probe and select Configure.
    • Infrastructure Manager: Expand the hus node and double-click the hub probe.
  2. Navigate to Queue List or Queue.
  3. If you use attach/get queues, set up the queues with the corresponding values:
    • Attach queue on the downstream hub:
      • Active: enabled
      • Name: attachDiscovery (or other name of your choice)
      • Type: attach
      • Subject: probe_discovery
    • Get queue on the upstream hub:
      • Active: enabled
      • Name: getDiscovery (or other name of your choice)
      • Type: get
      • Address: address of the hub that has the attach queue
      • Queue: name of the corresponding attach queue
      • Bulk size: number of messages to be transferred together. (Optional: if you expect the queue to carry a significant number of messages, send them in bulk to improve performance.)
  4. If you use post queues, set up each queue with the corresponding values:
    • Post queue on the downstream hub:
      • Active: enabled
      • Name: postDiscovery (or other name of your choice)
      • Type: post
      • Subject: probe_discovery
      • Address: address of the upstream hub
      • Bulk size: number of messages to be transferred together
  5. Repeat the previous steps on all hubs that require a queue.
Tip: In small to medium deployments, a wildcard (*) subject, which carries any message, can simplify queue configuration. Use of a wildcard subject in large installations is not recommended.

For queue setup details, click the question mark or Help button in the configuration GUI.

The following illustration shows discovery queue configuration using attach/get pairs.  If you use a post queue configuration, the flow in the illustration would be similar, but you set up post queues instead of attach queues on all downstream hubs and you would not need to set up get queues.

HUB--Setting up Queues (2)

When you have set up all required queues, run an automated discovery scan to confirm that the queues are operational. Review the list of discovered devices. In addition to local devices, this list would also contain devices that are only addressable from the secondary hubs.

Note: Setting up other queues for alarms, QoS, and baseline data is a similar procedure of configuring attach and get queues. The subject of the queue changes as required by the type of data to be carried.

Was this helpful?

Please log in to post comments.