Agents page

2 minute read

The Agents page lets you view all configured agents. If an agent is missing, check the hardware for proper connection and setup and check that the agent is assigned to the correct Cluster Manager host name.

Functionality

  • Click an agent name in the Name column to open the Agent Details page, where more information is available for that specific agent.

  • Click a host in the Host column to filter the agent list to agents on that host only.

  • Click a build in the Build column to open that build’s Build Details page.

  • Click an agent in the Enabled column to enable or disable that agent.

  • Click the test agent (test agent) button to test the agent in that row. For details, see Testing Agents.

  • After filtering agents , you can then enable , disable , test , or delete the agents in the filtered results by clicking the enable agent , disable agent , test agent , or stop agent button respectively.

  • The Actions column lets you view agent-related messages or delete the agent in that row

  • You can also:

  • Click a column heading to sort the information in that column

  • Set the page refresh interval to keep the information current

  • Set the number of records you want to see per page

Column descriptions

Column Description

Name

Name of the agent, which is a combination of the host name and a sequential, Cluster Manager assigned number.

Host

DNS host name.

Platform

Platform on which the agent is running.

Agent Version

Release number of software installed on the host. CloudBees recommends that the release number match the Cluster Manager release number.

EFS Version

Release number of the EFS version that is installed on the host.

Build

Name of the build that the agent is currently servicing.

Enabled

Whether the agent is available for use. If the agent is enabled and alive, the Cluster Manager can use the agent.

Status

checkmark Agent is alive (is running and connected to the Cluster Manager).

error Agent is not alive (cannot connect to the Cluster Manager).

penalty Agent failed and was not used for the build. The build will attempt to re-use the agent for later builds. If a build job causes an agent failure, that agent is placed in “penalty” status for 30 seconds. This prevents the agent from being reassigned to another build job during that time. The intent is to prevent an agent that has a fatal hardware failure from being repeatedly reassigned to build jobs and then causing multiple (successive) failures, which would eventually cause the entire build to fail.

Actions

view agent messages Opens the Messages page for viewing all system log entries for the agent.

delete Deletes the agent.

test agent Tests the agent.