# Netflix/concurrency-limits

Repository: https://github.com/Netflix/concurrency-limits
Canonical: https://ross.abutalabs.com/products/concurrency-limits
Language: Java
License: Apache-2.0
License Family: permissive
Last push: 2026-01-16T00:05:22+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 62, release rhythm 28, longevity 100
- inputs: {"age_days": 3187, "days_push": 230, "days_rel": 268, "gap_med": 229.5, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3598, forks 327 (observed 2026-08-28T04:08:11.388696+00:00)

## What it is
A Java library that applies TCP congestion control concepts to automatically detect and enforce concurrency limits for services. It uses latency-based algorithms like Vegas and Gradient2 to adapt limits as systems scale, achieving optimal throughput and latency.

## Use cases
- auto-detect concurrency limits for microservices
- gracefully shed excess load instead of falling over
- replace static RPS-based rate limits with adaptive concurrency limits
- prevent queue buildup and latency spikes under bursty traffic
- enforce per-service concurrency limits in auto-scaling distributed systems

## When to choose
- you run auto-scaling services where static RPS limits go stale
- you need latency-aware adaptive load shedding in Java/JVM services
- you want congestion-control-inspired backoff using timeouts and rejected requests

## When to avoid
- you need simple static token-bucket rate limiting
- you are not on the JVM and cannot port the concepts
- you need distributed global rate limiting across clusters rather than per-node limits

## Facets
- artifact type: library
- maturity: active
- function: rate-limiting, concurrency, middleware
- domain: microservices, performance, backend
- platform: jvm
- tags: concurrency-limits, load-shedding, tcp-congestion-control, adaptive-limits, latency-based

## Member repositories
- Netflix/concurrency-limits (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:11.388696+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:33:46.535057+00:00, confidence not recorded.
  - readme: https://github.com/Netflix/concurrency-limits (fetched 2026-08-28T04:08:11.388696+00:00, sha 37c086e5b2ba)
- Data as of 2026-08-30T08:39:29.467469+00:00.