Resource Auto-Discovery Processes

Topics marked with * relate to features available only in vFabric Hyperic.

This page describes VMware vFabric™ Hyperic® processes for discovering resources running on a platform.

Hyperic Auto-Discovery Processes

Most resources are automatically discovered by the Hyperic Agent running on a platform. The agent scans a platform to discover new resources, and resources whose properties have changed since the last scan. For instance, if a platform's IP address changes, the next scan detects the change.

Under the hood, an agent's auto-discovery capabilities break down to three different types of scan, described in the sections that follow.

Default Scan

A default scan discovers platforms and servers from the process table or Windows registry.

On Unix-like platforms, a default scan checks the process table for processes that match a given pattern. On Windows platforms, a simple registry scan is performed, looking for registry keys that installed products register during their installation process.

A default scan is performed upon agent startup, and every 15 minutes thereafter. A default scan typically does not take long and is not CPU-intensive.

You can also initiate a default scan on-demand for a platform to discover new servers. When you initiate a default scan, you can initiate a file scan, described below, as well.

On-Demand Default Scan Discovers all vSphere Resource Types

When you initiate a default scan on a platform where a vSphere vCenter server runs, it discovers any new vCenter instances, and als discovers all of the ESX hosts and VMs a vCenter instance manages.

images/download/attachments/79038211/DefaultScan.png

File Scan

A file scan discovers servers by scanning the platform's file system for manageable products' installation directories. You can configure what server types to look for and directories to include or exclude for the search. You can initiate a file scan explicitly; when you run a default scan, you can start a file scan at the same time — the agent never runs a file scan automatically.

images/download/attachments/79038211/manualscan.png

Runtime Scan

A runtime scan discovers servers and services — platform services as well as services that run in a server. The agent performs a runtime scan when a new platform or new servers on a platform are added to inventory. In addition, the agent automatically does a runtime scan once a day.

For information about how to scan a platform on demand and options for configuring and disabling auto-discovery behavior, see Options for Running and Controlling Resource Discovery in vFabric Hyperic Administration.

Dynamic Service Type Detection

The auto-discovery functionality described on this page is governed by the resource plugin that manages a resource type. In addition, the Hyperic Agent can auto-discover and manage Java application services via Model MBeans that adhere to a specified ObjectName naming convention and expose a specified set of service data. This enables you to monitor application services along with the hosting application server and its internal services. For more information, see Java Applications in vFabric Hyperic Resource Configuration and Metrics.

images/download/attachments/79038211/RuntimeScan.png

What the Agent Can and Cannot Discover

The data the Hyperic Agent discovers for a resource type is specified in the XML descriptor for the resource plugin that manages it. All operating system platform types are managed by HQ's system plugin. Most other Hyperic plugins discover multiple versions of a server type and the service types it contains. For example, Hyperic's tomcat plugin manages several versions of the Tomcat server and multiple services within Tomcat.

The plugins that ship with Hyperic are in the SERVER_HOME/hq-engine/hq-server/webapps/ROOT/WEB-INF/hq-plugins directory. If you want to monitor a software product for which Hyperic does not provide a plugin, you can build your own, using support classes available from Hyperic, and deploy it to this directory and to the Hyperic Agent.

There are some resources that the Hyperic Agent cannot discover:

  • Non-running resources - The agent cannot discover a resource that is not currently running.

  • Undiscoverable types - There are a few resource types that the Hyperic Agent can manage but not discover, even on the local platform. For example, you may need to configure the location of some versions of WebLogic Server. For a resource's configuration requirements, see the Configuration help section of the documentation for the resource plugin on HyperForge, the Hyperic community site.

How Discovered Resources Get into Hyperic Inventory

After a scan is completed, new and changed platforms and servers appear in the Auto-Discovery portlet on the Hyperic Dashboard. You explicitly add new and changed platforms to inventory - using controls in the Auto-Discovery portlet, or the Auto-Discovery Results page, which provides mored detailed results of what a scan detected on a platform. New and changed services do not appear in the Auto-Discovery portlet or the the Auto-Discovery Results page — you do not need to explicitly add them to inventory. When you add a platform or server to inventory, the associated services are automatically added as well.

What to Do After Adding New Resources to Inventory

After adding a new platform or server to inventory, you might need to:

  • Configure the resource to enable monitoring. Configuration options are found on the resource's Inventory page.

  • Add the resource to one or more new or existing resource groups. You can select groups to assign the resource on its Inventory page. Groups are useful for:

    • Resource access control, if you use vFabric Hyperic.

    • Monitoring a group of resources of the same type.

  • Add new services to new or existing applications to enable monitoring at the application level.