Keeping remote devices up to date is one of the hardest problems in fleet management. You cannot SSH into every node. You cannot assume someone at the remote site will run an update command. And you definitely do not want to ship a technician to every location for a software update.
How Auto-Update Works
ProxyPass handles this with auto-update on connect. It is a per-group setting that you toggle in the dashboard. When enabled, every node that connects — or reconnects — to the cloud is automatically checked against the latest available version. If the node is outdated, it receives an update command through the tunnel and upgrades itself.
On Linux, the update process uses a co-container approach. The running node spawns a lightweight helper container that downloads the new version and replaces the old one. The node goes offline briefly during the swap and reconnects with the updated version. On Windows, the native service handles the update through its own self-update mechanism.
Scenario: The Node That Was Offline for Three Months
A construction company deploys ProxyPass nodes on monitoring equipment at building sites. During winter, a site shuts down for three months. The equipment is powered off. You release two updates during that time — a performance improvement and a security patch.
Spring comes. The site reopens. The equipment powers on. The ProxyPass node reconnects to the cloud. Because auto-update on connect is enabled, the node immediately receives the latest version — skipping directly to the current release. Within a minute, the node is online, fully updated, without anyone touching it.
If auto-update had not been enabled, that node would run a version three months old until someone manually triggered an update. With auto-update, your entire fleet converges to the latest version automatically, even nodes that were offline for extended periods.
Manual Control When You Need It
You can also trigger updates manually for specific nodes through the Management API or the dashboard if you prefer a controlled rollout. And you can disable auto-update on connect for groups where you need to test new versions before deploying them to production nodes.
Full Lifecycle Control
The combination of auto-update and remote restart (which lets you reboot the container or the host OS through the tunnel) means you have full lifecycle control over every node in your fleet — without ever needing direct access to the machine.
No SSH. No physical access. No technician dispatched. Your fleet stays current.