The Jaws Deploy agent is a background service that runs on each machine you deploy to. It connects outbound to Jaws Deploy Cloud or your Stack server, receives deployment work, runs it on the machine and streams the log back.
Install steps
- In the workspace, open Machines and choose Add machine.
- Download the installation script for Windows (
jaws_agent_install.ps1, PowerShell) or Linux (jaws_agent_install.sh, bash). - Run the script on the machine. It needs no input and logs its progress to the console.
- The agent connects and sends a handshake. An administrator approves it in the workspace, and from then on the machine can receive deployments.
The script belongs to the workspace you downloaded it from. To add a machine to another workspace, download the script from that workspace.
Machines without internet access
The installer downloads the PowerShell modules it needs (PackageManagement 1.4.0 or later, and Microsoft.PowerShell.Archive). On an air-gapped machine, save those modules into a zip on a connected machine, copy the zip across, and run the script in offline mode:
# Windows
.\jaws_agent_install.ps1 -localPsModulesZip "C:\temp\jaws_ps_modules_offline.zip" -offlinePsModules $true
# Linux
bash jaws_agent_install.sh -m /tmp/jaws_ps_modules_offline.zip -o true
Network
The agent opens an outbound connection to Jaws Deploy and never accepts inbound connections, so the machine needs no inbound ports.
The agent needs outbound HTTPS (port 443) to app.jawsdeploy.net, which it uses during installation, and to the hub it connects to, currently hub-001.jawsdeploy.net. New hubs follow the same pattern (hub-002.jawsdeploy.net and so on). If your firewall allows rules by domain, allowing *.jawsdeploy.net covers future hubs. Stack agents only need your own Stack server.