blog/Backpressure & Load Control

Topic · 5 posts

Backpressure & Load Control

Runnable Java models of congestion collapse, admission control, token and leaky buckets, load shedding, and SLO-driven load control. Deterministic outputs, checked-in CSVs, reproducible on any machine.

posts5

first post2026-07-19

latest2026-08-16

total read~30 min

All Project Loom Structured Concurrency Tail Latency & System Behavior

5 posts

newest first

012026-08-16
Bounded Systems Architecture and SLO-Driven Load Control
Priority admission keeps 99.7% critical success at 300 rps; blind shedding delivers 32.3%. Background pays with 12.9%.
6 min
022026-08-16
Load Shedding Strategies
Three shedding policies, the same 100 rps goodput at twice capacity. Unbounded FIFO falls to 7.8 rps.
6 min
032026-08-16
Token Bucket vs Leaky Bucket
Two gates, near-identical goodput at 47.8 and 48.0 rps. The downstream peak differs: 290 rps against 100 rps.
6 min
042026-08-02
Admission Control Design
A deterministic sweep: goodput peaks at an admission limit of capacity times deadline, then collapses eight slots later.
6 min
052026-07-19
Why Systems Collapse Under Load
110 rps against 100 rps of capacity drops goodput to 42 rps, and retries make it worse. A deterministic Java model.
6 min