Day 24 results: partitions, consumer groups, and the ordering guarantee

Measured on: (machine, OS, Python version)

Output

Paste part 1, part 2, part 3 and the scoreboard here.

One sentence per prediction

P1 (how many partitions one key’s records land on):
P2 (throughput speedup at C = P consumers):
P3 (idle consumers at C = 2P):
P4 (per-key order violations):

Partitioning, in my own words

Why does one key always go to one partition, and why do many different keys spread across all of them:

Consumer groups: where the line goes flat

Throughput climbed to about ___x at C = P, then stopped. How many consumers sat idle at C = 2P, and why can you never have more useful consumers than partitions:

The ordering guarantee people get wrong

Within a partition the out-of-order count was __, across two partitions it was __, and per key it was ___. Say the guarantee Kafka actually gives in one sentence:

The number that matters

Parallelism is capped at the partition count, and order holds only inside a partition (so pick your key to match the thing you need ordered):