It has been known for some time that we would benefit from a profiling/optimizing deep-dive for Flow-IPC (with a likely focus on SHM-jemalloc which has shown signs of adding significant latency and/or processor cycle use, informally). I noted this near the top of the official docs. This is the ticket for that.
It has been known for some time that we would benefit from a profiling/optimizing deep-dive for Flow-IPC (with a likely focus on SHM-jemalloc which has shown signs of adding significant latency and/or processor cycle use, informally). I noted this near the top of the official docs. This is the ticket for that.