Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2508.16119 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916912003284992 |
|---|---|
| author | Gaikwad, Madhava Gandhi, Abhishek |
| author_facet | Gaikwad, Madhava Gandhi, Abhishek |
| contents | We present ANSC, a probabilistic capacity health scoring framework for hyperscale datacenter fabrics. While existing alerting systems detect individual device or link failures, they do not capture the aggregate risk of cascading capacity shortfalls. ANSC provides a color-coded scoring system that indicates the urgency of issues \emph{not solely by current impact, but by the probability of imminent capacity violations}. Our system accounts for both current residual capacity and the probability of additional failures, normalized at datacenter and regional level. We demonstrate that ANSC enables operators to prioritize remediation across more than 400 datacenters and 60 regions, reducing noise and aligning SRE focus on the most critical risks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_16119 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ANSC: Probabilistic Capacity Health Scoring for Datacenter-Scale Reliability Gaikwad, Madhava Gandhi, Abhishek Networking and Internet Architecture Artificial Intelligence 60K20 (Primary), 90B25, 68M15 (Secondary) C.2.4; D.4.7; I.2.6 We present ANSC, a probabilistic capacity health scoring framework for hyperscale datacenter fabrics. While existing alerting systems detect individual device or link failures, they do not capture the aggregate risk of cascading capacity shortfalls. ANSC provides a color-coded scoring system that indicates the urgency of issues \emph{not solely by current impact, but by the probability of imminent capacity violations}. Our system accounts for both current residual capacity and the probability of additional failures, normalized at datacenter and regional level. We demonstrate that ANSC enables operators to prioritize remediation across more than 400 datacenters and 60 regions, reducing noise and aligning SRE focus on the most critical risks. |
| title | ANSC: Probabilistic Capacity Health Scoring for Datacenter-Scale Reliability |
| topic | Networking and Internet Architecture Artificial Intelligence 60K20 (Primary), 90B25, 68M15 (Secondary) C.2.4; D.4.7; I.2.6 |
| url | https://arxiv.org/abs/2508.16119 |