Datacenter RPCs can be General and Fast

TL;DR

eRPC demonstrates that general-purpose RPC libraries can achieve high performance comparable to specialized systems on commodity hardware, supporting scalable, low-latency datacenter communication.

Contribution

This paper introduces eRPC, a general-purpose RPC library that matches the performance of specialized systems on commodity datacenter hardware without requiring niche technologies.

Findings

Handles up to 10 million small RPCs per second per CPU core

Achieves 75 Gbps bandwidth for large messages

Provides 5.5 microseconds latency for Raft replication on lossy Ethernet

Abstract

It is commonly believed that datacenter networking software must sacrifice generality to attain high performance. The popularity of specialized distributed systems designed specifically for niche technologies such as RDMA, lossless networks, FPGAs, and programmable switches testifies to this belief. In this paper, we show that such specialization is not necessary. eRPC is a new general-purpose remote procedure call (RPC) library that offers performance comparable to specialized systems, while running on commodity CPUs in traditional datacenter networks based on either lossy Ethernet or lossless fabrics. eRPC performs well in three key metrics: message rate for small messages; bandwidth for large messages; and scalability to a large number of nodes and CPU cores. It handles packet loss, congestion, and background request execution. In microbenchmarks, one CPU core can handle up to 10 million small RPCs per second, or send large messages at 75 Gbps. We port a production-grade implementation of Raft state machine replication to eRPC without modifying the core Raft source code. We achieve 5.5 microseconds of replication latency on lossy Ethernet, which is faster than or comparable to specialized replication systems that use programmable switches, FPGAs, or RDMA.