i have a node-red instance running on kubernetes which works really fine in general.
On that instance i'm controlling some media services for tv-broadcasting and one of it is CasparCG which uses AMCP protocol. Therefor i'm using the telnet-send node to send commands to that CasparCG.
Last days i had a connection issue to that CasparCG - firewall-rules made a problem - and Node-Red coudn't send commands to that telnet-endpoint.
In consequence node-red crashed and restartet.
Message was only "Connection wait exceeded timeout"
Is there a way to secure this node from crashing the whole instance?
In general, a node should not crash Node-RED. Node-RED does its best to handle errors but some (like unhandled async errors) cannot (and should not) be handled by Node-RED but rather It is down to the node creator to handle async errors.
Which node are you using (like to library) - perhaps there is a link to its repository and it is well maintained - the author may be willing to fix it up or accept a pull request?