[In Progress]Module_ID: #04_OBS

Web Observability Dashboard

A live control room for a high-traffic API — streaming metrics, traces, and alert state onto one pixel-perfect grid.

Project Overview

This dashboard is the single pane of glass for a high-concurrency API. A Go collector fans out Prometheus scrapes, a rule engine turns plain-language thresholds into alerting expressions, and a React front end renders the whole system state on a retro grid. Trace drill-down ties every latency spike back to the request that caused it. (In active development.)

System Features

Live metrics: Prometheus scrapes stream into a React grid refreshed on a sub-second tick.
Custom alerts: A rule builder compiles thresholds into Prometheus alerting expressions.
Trace drill-down: Click any spike to open the correlated request trace.

SYS_SPECS

Frontend:: React
Metrics:: Prometheus
Dashboards:: Grafana
Collector:: Go

ARC_MAP

[Metrics Pipeline]

Challenges & Solutions

# Threat: Cardinality Explosion

High-dimensional labels ballooned the metric series count until dashboard queries timed out.

# Remedy: Label Budgets

A recording-rule layer pre-aggregates hot queries and caps label cardinality at ingest.