SRE Your gRPC...
summary
Shows concrete changes to a gRPC C++ stack: use Protocol Buffers for compatible APIs, add a translator backend, set per-RPC deadlines via flags, and propagate remaining deadline from server to downstream services. Recommends simple load spreading: round‑robin or deterministic shuffles over full-mesh, prefer wait_for_ready=false, return cheap defaults or cached translations on timeout, and log counts of deadline/errors. Claims these small edits avoid cascading failures and make latency and capacity behavior observable.
AI-written summary · about
chapters (9)
- 0:00Start
- 10:41adding the translators to the graders
- 26:39set the deadlines
- 33:19specify the longest deadline
- 33:55specify the longest default deadline
- 40:29log and error
- 45:05thinking about the lifecycle of your tasks
- 47:18setting a default deadline for your service
- 50:20compute your performance
related talks
48mSRE Your gRPC—Building Reliable Distributed Systems Illustrated with gRPC
35mLooking at SRE Needs and Trends over Two Decades with a...
33mIntro to gRPC
21mR2P2: Making RPCs first-class datacenter citizens
32mgRPC Deep Dive: Prevent Your Service From Overtaking Itself
35mEvolving Critical Systems With gRPC: Bidirectional at Scale — What Works, What Hurts, What’s Next?
15mRemote Procedure Call as a Managed System Service
01h 42mWhat is a Reactive Application?