Build Monitoring Dashboard & Alerting Policies for strr-api

Objective

Establish comprehensive performance visibility and automated error alerts for the strr-api Cloud Run service using GCP Cloud Monitoring. This will ensure real-time tracking of endpoint health and minimize response time for production operational failures.

Tasks & Requirements

1. Configure Cloud Monitoring Dashboards

Build a unified telemetry dashboard using Prometheus/OpenTelemetry metrics or custom log metrics extracted from Cloud Run request logs, capturing the following:

• 1a. Latencies per Endpoint: Extract request latency fields to chart the 50th, 90th, and 99th percentiles. Metrics must be cleanly grouped by HttpRequest.requestUrl or HTTP route templates (e.g., POST /applications).


• 1b. Errors per Endpoint: Visualize error distribution grouped by endpoint route, filtering request volumes to flag client-side errors (HTTP status codes 4xx) and server-side errors (HTTP status codes 5xx).

2. Implement HTTP Code-Based Alerting

Metric / Scope | Trigger Condition | Evaluation Window
-- | -- | --
HTTP 4xx Client Errors (Per-Endpoint Route) | > 5% of total incoming traffic | 5-minute rolling window
HTTP 5xx Server Errors (Per-Endpoint Route) | > 1% of total incoming traffic | 5-minute rolling window

Start with the strr-api Cloud Run deployment and its request logs, then review how GCP Cloud Monitoring can expose latency and HTTP status metrics grouped by route. Done means a dashboard shows the requested latency percentiles and 4xx/5xx distributions, with 5-minute endpoint alerts at the stated thresholds.

Written by the indexing model from the issue text.