mmukhi

proto: add scatter/gather API for serialization

13

Serialized data read from the wire often spans several data frames. The current `proto.Unmarshal` API requires that data from all these frames be copied to a slice. This extra copying(and...

transport: reuse memory on read path

51

One possible idea is to use a leaky buffer.

Under 80% network utilization for LAN networks with large messages and high concurrency

2

go run benchmark/benchmain/main.go -benchtime=10s -workloads=streaming -compression=off -maxConcurrentCalls=1000 -trace=off -reqSizeBytes=1048576 -respSizeBytes=1048576 -networkMode=LAN -cpuProfile=speedup.cpu

Number of entries and size of each entry should be tweaked through settings. Ideally we could independently configure client headers, server headers, and server trailers. It would be interesting to...