snipercze / corona Goto Github PK

When nmap scan, corona server tries to reply to connected client but it fails (due nmap is already done). Corona server then crash with message:

Traceback (most recent call last):
File "./corona-server.py", line 12, in
serverSocket.bind((host, int(port)))
OSError: [Errno 98] Address already in use

Agent is trackable only when active right know. When agent goes offline, server forget any information about it and agent is reregistered as brand new node when appear online. This should be fixed.

Because of "cloudy" modern IT design we cannot use IP as identificator across time. So each agent should have some UUID generated by server, which is uniq and stored at agent side somewhere in /var/run...

New agent is connecting and no UUID assigned - server generate new UUID, save it with agent data to database and send this UUID back to agent. Agent will save it.
Agent is reappearing and has UUID assigned - agent sends it to server and server just mark already existing agent as online (based on UUID). When UUID exists but IP and/or port of agent is not corresponding to database settings, this agent is rejected.
Agent is disappearing/healthcheck fails - server mark this agent (identified by UUID) as offline.

When agent is configured to connect to server which is unreachable, agent crasher with ConnectionRefused exceptions