About Agent - Server Communication

This section is an overview of the communications between a Hyperic Agent and the Hyperic Server.

Agent-to-Server Communication

Upon startup, a Hyperic Agent initiates a communications channel with the Hyperic Server. The agent continuously batches and sends monitoring results for the platform to the server. Agent-to-server data flows over HTTPS via a byte-encrypted XML protocol called Lather.

The agent sends this data to the server:

  • SSL certificate — When an agent connects to a server using https, it presents an SSL certificate as part of the handshake. For information about certificate verification, see Hyperic 4.6.x SSL Certificate Processing on About SSL in Hyperic.

  • plugin status — When an agent starts up and loads the plugins in its plugin directory, it sends a plugin status report to the server, including the MD5 checksum of each plugin it loaded. This server uses this information for the server-agent plugin synchronization (SAPS) process, described in About Plugin Sync at Agent Startup on Plugin Deployment and Management in vFabric Hyperic Administration.

  • metrics — You can configure the batch size is using the agent.maxBatchSize property in agent.properties

  • event — You can configure the batch size using agent.eventReportBatchSize in agent.properties

  • auto-discovery results — Auto-discovery results are reported after each auto-discovery scan.

    • By default, the default scan runs every 15 minutes. You can configure the frequency using the autoinventory.defaultScan.interval.millis in agent.properties

    • By default, the run-time scan runs once a day. You can configure the frequency using the autoinventory.runtimeScan.interval.millis in agent.properties.

Server-to-Agent Communication

Unless a Hyperic Agent is configured for unidirectional communication, the Hyperic Server initiates communication with a Hyperic Agent to provide metric collection schedules, and relay the commands and data issued or configured by authorized users.

Server-to-agent traffic is always over SSL, and includes:

Unidirectional Agent-Server Communications

The default communication between an agent and the server is bi-directional — the Hyperic Agent establishes a connection with the Hyperic Server to send the data described in Agent-to-Server Communication, and the Hyperic Server establishes a connection with the Hyperic Agent to send the data described in Server-to-Agent Communication.

If your security policies dictate, you can configure the agent to initiate all communications with the Hyperic Server. You can configure unidirectional communications interactively, at first agent startup, or in an agent's AgentHome/conf/agent.properties file. If you want to change from bidirectional to unidirectional communications after initial agent configuration, see Configure Unidirectional Agent - Server Communication.

If an agent is configured for unidirectional communication, the Hyperic Server does not establish a connection to the agent. Instead, at startup, the agent establishes a persistent connection to the Hyperic Server, and polls the server periodically for server communication.

Unidirectional Communication via a Proxy

If you use unidirectional communications, you can configure the Hyperic Agent to communicate with the Hyperic Server via a proxy server. You must configure this behavior in each agent's agent.properties file. For more information, see Set Up Agent in Properties File.