HTTP/2 lets gRPC run many calls at once over one connection, in a compact binary format.
Multiplexing sends many requests and responses at the same time over a single connection.
Binary framing is cheaper to parse than the text protocol of HTTP/1.1.
Header compression with HPACK cuts the bytes spent on repeated headers.
Flow control stops a fast sender from overwhelming a slow receiver.
Reusing one connection avoids new handshakes and cuts latency.
HTTP/2 also defines server push, but gRPC doesn't use it. The grpc-java client turns push off in its HTTP/2 settings, for example. Server streaming covers the case of many responses to one request.
Rewriting in plainer words…
This answer doesn't lend itself to a diagram - it reads best . No credits were charged.
Saved in this browser - sign in to keep your review list.
How should your speech become text?
Listening… your words appear above as you speak - tap Stop when you're done.
Recording · cr - tap Stop & transcribe when you're done.
Transcribing with AI…
Voice:
Keep going - a few more words and AI can grade it.
Interview lens
How interviewers actually use this question
Likely follow-ups
Multiplexing stops requests queuing behind each other. What happens to every stream when one TCP packet is lost?Name TCP head-of-line blocking. Every stream on the connection waits for the resend, and HTTP/3 over QUIC fixes this per stream.
HTTP/2 lets gRPC run many calls at once over one connection, and it sends them in a compact binary format instead of text.
Multiplexing matters most, because requests and responses run as separate streams on one connection and don't wait in line behind each other like they do on HTTP/1.1.
Binary framing is cheaper to parse than HTTP/1.1 text, and HPACK compresses headers, so the headers repeated on every call cost far fewer bytes.
Flow control stops a fast sender from overwhelming a slow receiver, and reusing one connection skips new handshakes, which cuts latency.
HTTP/2 also defines server push, but gRPC doesn't use it, and server streaming handles many responses to one request.
One downside is that a single long-lived connection sends every call to one backend behind an L4 load balancer, so gRPC needs per-call balancing at L7 or in the client.
Weak answers to avoid
Says gRPC uses server pushgRPC doesn't use HTTP/2 server push. Many responses to one request come from server streaming, so name that instead.
Lists features with no payoffReciting binary, HPACK and multiplexing is just a list. Tie each one to the cost it cuts, like parse time, header bytes, handshakes or calls waiting in line.
Claims multiplexing ends all head-of-line blockingIt stops HTTP requests queuing behind each other, but a lost TCP packet still stalls every stream on the connection. Say that plainly if asked.
Credits HTTP/2 for Protobuf's small payloadsProtobuf makes the message body compact. HTTP/2's binary framing and HPACK shrink the framing and headers. Keep the two layers apart.
Interview lens
Likely follow-ups, what you can say, and the weak answers to avoid.
