Installation
Flopsar is installed in two parts: the server (which also serves the Workstation web application) and the agents (attached to your JVM applications). This chapter walks through both, plus the step that ties them together — issuing an agent token.
Before you begin, make sure that:
you have a 64-bit GNU/Linux host for the server, with a few gigabytes of free disk space for the collected data;
your network allows TCP connections from your application hosts to the server.
Installation Order
The order of the steps matters. Install Flopsar in the following sequence:
Install the server. Set up the Flopsar server, accept the EULA, and start the service. See Server.
Create an agent token. For each application you want to monitor, generate an agent token (a pre-shared key) on the server. See Agent Tokens.
Install the agents. Attach the agent to each JVM, configuring it with the application name, the server address, and the token from the previous step. See Agent.
You must complete these steps in order. An agent will only connect once the server is running and the agent is configured with a valid token for its application. If you start an agent before the server exists or without a matching token, the agent cannot authenticate — and the instance will not appear in the Workstation. Always install the server and create the agent token first, then install the agents.
What Gets Installed
Server — a single binary with an embedded HTTP server that serves the REST API and the Workstation. It listens for agent connections, stores the collected data, and enforces licensing. See Server.
Agent — a native library loaded into each monitored JVM. It connects to the server, applies the instrumentation profile it receives, and streams telemetry back. See Agent.
To run the server in a container instead of installing a package, see Docker. To send OpenTelemetry traces to the same server, see OpenTelemetry and the [otlp] section of the Server Configuration Reference.
Last updated