Aaron Stannard
Aaron Stannard
This changes DotNetty's threading mechanics in a significant way: https://github.com/Azure/DotNetty/commit/05e5aa40c3b1078ee80d520f3516bed7ebf11520 Worth benchmarking
Just benchmarked this - doesn't seem to be any improvements, and in fact might be worse than what we have in `dev`.
@notour would you like to send an updated PR with that code change? Appreciate the info!
@john-cullen we should probably implement @leo12chandu's changes
Definitely a bug.
> This way we can grow potential size of the payload while serializing, without paying the cost of copying the memory to bigger buffer. Ah ok, that's similar to how...
> The most critical aspect to this is the core write/read types, TBufferWriter : IBufferWriter & ReadOnlySequence. By landing on them we can reduce impedance mismatch. Agree - if JSON.NET,...
Related: https://github.com/akkadotnet/akka.net/pull/6026
Just ran a simple test so far - formed a cluster with this change no problem. Curious to see if the test suite explodes on contact or not.
If this escapes beta stage, I'm going to fork the `HashedWheelTimer` and include a copy that runs with the old behavior and set that to default in order to limit...