teleo-codex/domains/internet-finance/square-root-staffing-principle-provisions-servers-as-base-load-plus-beta-times-square-root-of-base-load-where-beta-is-quality-of-service-parameter.md
Teleo Pipeline e0c9323264
Some checks are pending
Sync Graph Data to teleo-app / sync (push) Waiting to run
extract: 2019-00-00-whitt-what-you-should-know-about-queueing-models
Pentagon-Agent: Ganymede <F99EBFA6-547B-4096-BEEA-1D59C3E4028A>
2026-03-15 17:12:39 +00:00

30 lines
No EOL
1.9 KiB
Markdown

---
type: claim
domain: internet-finance
description: "Optimal server provisioning follows R + β√R formula where R is base load and β controls service level"
confidence: proven
source: "Ward Whitt, What You Should Know About Queueing Models (2019)"
created: 2026-03-11
---
# Square-root staffing principle provisions servers as base load plus beta times square root of base load where beta is quality-of-service parameter
The square-root staffing rule provides optimal server provisioning: if base load requires R workers at full utilization, provision R + β√R workers where β ≈ 1-2 depending on target service level. This formula emerges from queueing theory analysis of multi-server systems and represents the sweet spot between over-provisioning (wasteful) and under-provisioning (queue explosion).
The principle applies across domains: call centers, compute pipelines, service systems. For Teleo pipeline scale (~8 sources/cycle, ~5 min service time), this gives concrete worker count guidance without requiring peak-load provisioning.
The underlying insight: variance in arrival and service times creates queueing delays even when average utilization is below 100%. The square-root safety margin handles this variance efficiently. The margin grows with system size but at a sublinear rate, creating economies of scale.
## Evidence
Ward Whitt's practitioner guide establishes this as the foundational staffing principle in operations research. The formula derives from the Halfin-Whitt heavy-traffic regime analysis, where systems operate near full utilization (approaching 1 at rate Θ(1/√n) as servers n grow) while keeping delays manageable.
Erlang C formula provides the computational implementation for determining β given target service levels (probability of delay, average wait time).
---
Relevant Notes:
- domains/internet-finance/_map
Topics:
- core/mechanisms/_map