You've seen what the agents are and when they can be used. However, how do we set up an agent and let it communicate with the master? Let's start with the second part of the question and describe the communication protocols between the master and the agent.
In order for the master and the agent to communicate, the bi-directional connection has to be established.
There are different options for how it can be initiated:
- SSH: The master connects to the slave using the standard SSH protocol. Jenkins has an SSH client built in, so the only requirement is the SSH daemon (sshd) server configured on slaves. This is the most convenient and stable method because it uses standard Unix mechanisms.
- Java web start: A Java application is started on each agent machine and the TCP connection is established between the Jenkins slave application and the master Java application. This method is often used if the agents are inside the fire-walled network and the master cannot...