On Fri, Apr 2, 2021 at 10:22 AM Weeks, Thomas <[email protected]> wrote:
> Forgive my ignorant first post.. but I hadn't seen anything posted on this > anywhere yet.. > > But we use a sharded Guac container config (w/external ha-proxy https > front end) running at large scale (thousands of AWS VMs).. and since > upgrading to v1.2, (preparing to upgrade to 1.3), but users are currently > seeing dropped RDP sessions (multiple times/day) on both Windows(10, server > 2016) as well as Linux (Ubuntu/Deb run xrdp). > > It physically looks like this: > https://photos.app.goo.gl/zo165UeDtQHoYWST6 > > and in the logs: > > 17:22:47.909 [Thread-271] ERROR o.a.g.w.GuacamoleWebSocketTunnelEndpoint - > Connection to guacd terminated abnormally: Connection to guacd timed out. > ... > 17:22:47.910 [Thread-271] INFO o.a.g.tunnel.TunnelRequestService - User > "f1bb5985-5f0a-4ce5-8d25-ac031d198c97" disconnected from connection > "a609c1f6-0b59-48e0-84eb-7850b62a6ddb/b7202e69-a7eb-4a68-8762-b60ca0815d68/2d5a891b-f7e7-4d99-b583-05376e5a2f5a". > Duration: 311400 milliseconds > > > Reconnecting (same session token) seems to work. > > Anyone seen this or know if it's related to a known issue? Never happened > on older guac (v 0.95) > The error displayed on the screen in your photo ("The Guacamole server is not currently reachable. Please check your network and try again.") indicates network disruption between the browser and the Guacamole server. The logs noting connection timeout also suggest that a network disruption resulted in a lack of communication from the client, with guacd ultimately closing the connection. The most likely cause of what you're seeing is a transient network failure or something between the browser and the server interfering with the connection, perhaps closing what it considers to be too long-running. Michael Jumper CEO, Lead Developer Glyptodon Inc <https://glyp.to/>.
