SocketCAN transport disconnections

[ottlukas]

SocketCAN transport can silently die due to swallowed exceptions:

2022-02-16 18:28:11.134 [WARN ] [io.netty.channel.DefaultChannelPipeline] - An exceptionCaught() event
was fired, and it reached at the tail of the pipeline. It usually means the last handler in the pipeline
did not handle the exception.
tel.schich.javacan.linux.LinuxNativeOperationException: Unable to read
from the socket - errorNumber=100, errorMessage='Network is down'
    at tel.schich.javacan.SocketCAN.read(Native
Method)
    at tel.schich.javacan.AbstractCanChannel.readSocket(AbstractCanChannel.java:160)
    at
tel.schich.javacan.RawCanChannelImpl.readUnsafe(RawCanChannelImpl.java:82) 
    at org.apache.plc4x.java.transport.socketcan.netty.SocketCANChannel.lambda$doConnect$0(SocketCANChannel.java:118)
 
  at java.lang.Thread.run(Thread.java:834) [?:?]

This error happened within an hour of application start so its most likely related to some CAN instability, however what's important is lack of proper handling of error. The connection in theory was still alive, however there was no data coming over it any more.

Imported from Jira PLC4X-334. Original Jira may contain additional context. Reported by: ldywicki.

[chrisdutz]

@splatch Is this still an issue?

[splatch]

I am afraid it is still an issue, I'd need to consult javacan if there is a way to spot shutdown of the interface.