I recently deployed an FRP-based remote connectivity platform on my website to help manage secure reverse proxy connections between remote Android devices and a centralized server environment. The platform initially worked without major issues, and tunnel creation, client authentication, and remote session management were all functioning normally. However, after running the service continuously for several days, I started noticing unstable tunnel behavior and inconsistent synchronization between the client devices and the web dashboard. Here is my website you can check https://gsmneofrpp.com.mx/

One of the main problems is that active FRP clients occasionally appear as offline on the dashboard even though the devices are still connected and responding in the background. In several cases, the server logs continue showing heartbeat packets from the client, but the website interface fails to refresh the status correctly. This creates confusion because the system reports inactive sessions while the reverse proxy tunnel is technically still operational and passing traffic normally.

Another issue involves delayed authentication responses during new client registrations. When a device attempts to establish a secure connection using token-based authentication, the request sometimes hangs for several seconds before finally connecting. During peak usage periods, a few devices fail authentication entirely and automatically retry multiple times. I checked the database queries and API responses, but there are no obvious errors being returned from the backend services.

I also noticed unusual CPU spikes and memory consumption on the FRP server whenever multiple encrypted tunnels are opened simultaneously. The web control panel becomes noticeably slower, especially while loading tunnel statistics and active traffic logs. In some cases, the browser session times out before the statistics page fully renders. Restarting the FRP service temporarily improves performance, but the issue eventually returns after several hours of uptime.

The SSL-enabled reverse proxy configuration may also be contributing to the instability. I configured HTTPS with automatic certificate renewal and custom subdomain routing for each connected client device. Since enabling SSL enforcement, some users report intermittent websocket disconnections and delayed remote responses. Static pages on the website load correctly, but real-time monitoring components occasionally freeze or stop updating until the page is manually refreshed.

I would appreciate suggestions from anyone who has deployed a similar FRP-based remote management platform at scale. I am particularly interested in recommendations related to optimizing tunnel persistence, reducing authentication latency, improving websocket stability, and handling high numbers of concurrent reverse proxy sessions more efficiently. If anyone has encountered similar synchronization issues between FRP services and a custom web dashboard, I would like to know what debugging methods or server-side optimizations helped resolve the problem.