On Thu, Apr 18, 2024 at 1:30 AM Viji Shankar <[email protected]> wrote:
> Hi Team, > > We are enabling auto-scaling for the Guacamole pod which contains 3 > containers(guacamole, guacd and ngnix). After giving load to the container, > the pod got autoscaled but we are getting the following error when we > connect VM using guacamole. > > > > ERROR: *An error has occurred and this action cannot be completed. If the > problem persists, please notify your system administrator or check your > system logs.* > > > You'll need to check the logs further, and possibly enable some additional debugging, to see what's going on, here. Look at container logs for both guacamole and guacd containers. > Please give me some suggestions to enable pod auto-scaling for guacamole. > Because Guacamole currently does not have any built-in HA capabilities (for synchronizing session and such), in a configuration like this you will need to insure that: * The front-end load balancer uses some sort of session tracking or "stickiness" to make sure that the client gets sent to the same Guacamole Client (guacamole) container. Otherwise, the client will log in to one of the guacamole containers, but then on the next request, potentially be redirected to another guacamole container, which won't "know" about the login. * The guacamole containers need to be configured to use a consistent guacd container for the connection. If any load balancing is configured such that it might send traffic to multiple back-end guacd containers, this will likely cause problems and result in errors. -Nick
