[M1] SSH panic in integration environment

[ivcosla]

Running the integration tests with docker environment the following panic happens (long).

ssh-panic.txt

[ivcosla]

During ssh integration testing I have been able to stop two different scenarios. I will describe both of them just in case one re-appears, however, I have been unable to reproduce case 1 again despite following the same integration process on each build.

On the first scenario, when using the following command therealssh-cli $PK_A hangs as you hit enter. This shows in visor-A that there is an attempt to use a nil rule, and it will show unkown transport ID. Such error is forwarded from this line: https://github.com/skycoin/skywire/blob/mainnet-milestone2/pkg/router/router.go#L212

On the second scenario therealssh-cli hangs after hitting enter and entering in what should be an ssh session. However, in the new session the cursor appears frozen. At this point in time, there won't be error logs on any of the visors. If you try to type something on the session it won't show it, however, each key you stroke is being send as a different package from visor C to A, and you can see this on the logs if you print each package payload.

On this last scenario I haven't found any error with the regular communication process. Connections are successfully established and everything seems to work correctly. My latest assumption, which is yet untested is that such error might be here https://github.com/skycoin/skywire/blob/mainnet-milestone2/cmd/therealssh-cli/commands/root.go#L113, in the two operations that copy input and output from the client.