rns.recipes

Could just be me but this erorr along with an EPOLLOUT error show up repeatedly in my log and just get more frequent the longer my node is up, anyone else?

Yes, a lot of people have been experiencing that since March. It appears to be because rnsd is overloaded with path requests and announces. I've been out of the loop lately, but I think 1.1.5 which was released yesterday attempts to fix it.

Thanks for the reply, can confirm that 1.1.5 has not fixed it but the issue i had with rnstatus becomeing unresponsive seems resovled touch wood.

The "no interfaces could process the outbound packet" is actually more of a warning. It happens if a packet was in-flight but then the interface disappeared, for example a TCP/backbone client disconnects while a transfer was in progress. I adjusted the loglevel to be more sensible in the latest release.

That being said, there was also a bug in the EPOLL modification error handling, which was also fixed.

Regarding the path resolution stuff, the 1.1.6 pre-release should fix it, but there's some pretty massive refactoring of the transport I/O core, so any help testing it out will be very useful. Should see a significant effect once a few of the large backbone nodes are updated.

But it really is a pre-release. There's a non-zero chance that it could contain threading deadlocks, since a lot of stuff was changed around. So if you try it out, keep an eye on it, and let me know if it deadlocks or anything else weird happens.

Thanks Mark, will try 1.1.6 and see how I go. I also observed "AES_128_CBC dropped" in the logs for the first time today too, guess there are still older nodes out there.