Timeout handling for /bin/bash /health-check.sh is missing about fhem-docker HOT 6 CLOSED

Actually it once was nc or telnet at the very beginning IIRC. However, the way fhem.pl handles i/o was not 100% compatible (i remember something about new lines and carriage return as well as general encoding / UTF-8). This is why this method uses the very own and official "API" that fhem.pl provides. You may even find some heated discussion about this from that time in the forum. The overall intention here is to minimize the varieties of potential issues and to "charm" fhem.pl as good as possible.

from fhem-docker.

We now have a timeout for fhem clientmode:

https://forum.fhem.de/index.php/topic,117250.0.html

But i think implementing #30 is also worth, because the script can also hang on othere lines for some time.

from fhem-docker.

I'd rather prefer this to be fixed in fhem.pl as timeout handling should be done where connectivity is handled. Other code relying on this connection method would also benefit from it.

That being said, it might be challenging to convince the fhem.pl maintainer to implement such change or even accept a pre-defined patch (based on experiences from the past). Good luck.

from fhem-docker.

I understand what you mean. May i have a chance.

What about using a more lightweight client to query fhem?

Calling fhem.pl every intervaltime seems a little bit wasty of ressources. I think a telnet or netcat command woud bei enough to get the response.

from fhem-docker.

I know these problems on the telnet api. Because it was so ugly i do all stuff via http / https requests on fhemweb.
This is much better to handle.

from fhem-docker.

Great! Thanks for taking the lead here. Let's continue discussing in PR #30 and consider this issue to be resolved for now.

from fhem-docker.