Hi @fibu-freak, I can definitely say that your rules are not the issue - the binding breaks occasionally, all by itself - I do not have any rules set up that involve the avmfritz binding at the moment.
What I can also say (just had the problem again, today): when the error occurs, there is no more http conversation between openhab and the fritz box on the network. Meaning, the issue is on the openhab side (as it’s openhab that is cyclically requesting device info data from the fritz box).
I’ll look into the exception message again, now - there was something I found a couple of weeks back wrt this, stating that it’s related to some issue in java libraries or something…
UPDATE: Allright, I’ve not much to go on, but the current hypothesis (based onthis google groups thread) is that the following happens in my case: occasionally, my fritzbox slows down to a crawl, indicated by the fact that network (Wifi) routing performance tanks, logging into/working with the web interface becomes unbelievably slow, etc - this either goes away by itself after a while, or it can be fixed by rebooting the box*.
I had one of those slowdowns this morning - and when I checked later today, the avmfritz binding exhibited the problem this thread is about.
So I’m guessing that during these slowdowns, the connection pool buffer use by openhab fills up, and never recovers.
If that is the case, a workaround could be to increase the pool size numbers; even better would be to find a way to recover from the situation. No idea whether that’s possible, but I’ll have a look around.
(*) Why the fritzbox slows down so much is also a good question; I haven’t investigated this yet, one assumption is that it’s to do with USB memory that I’ve mounted, and maybe it’s running an indexer on that stick occasionally, causing the slowdowns? No clue; I’ve removed the mounted storage now, and will look for information about this as well.