Production Continuous Profiling with eBPF and Pyroscope on Linux
Pinpoint CPU hotspots and memory allocations across production Linux processes without code modifications using eBPF probes and Grafana Pyroscope.
Observability8 min read
NetBay journal
Observability
15 practical Observability articles from Netbay Engineering — written for operators running Linux and Windows VPS in India.
15 articles in this topic
AllLinuxLinux AdministrationWindowsLinux SecurityLinux NetworkingContainersContainers & OrchestrationKubernetesCI/CDGitHubSelf-HostingBackups & DataDatabasesApp DeploymentDNS & TLSObservabilityEmailAI AgentsAI InfrastructureAPI & AutomationProxiesCloud ArchitectureHardware & PerformanceDevOps Practices
Scaling Metrics Storage with VictoriaMetrics: 10x Compression Over Prometheus
Deploy single-node VictoriaMetrics on Ubuntu to store years of high-resolution Prometheus metrics with sub-second queries and minimal disk usage.
Observability7 min read
Real User Monitoring (RUM) with Grafana Faro and OpenTelemetry
Capture real-world user performance, Core Web Vitals, and JavaScript exceptions using the Grafana Faro Web SDK connected to an OTel Collector.
Observability6 min read
Advanced PromQL Patterns: SLI/SLO Calculations and Burn Rate Alerting
Implement production-grade Service Level Objectives and multi-window burn rate alerts using advanced PromQL queries on Prometheus 3.0+.
Observability7 min read
Architecting an OpenTelemetry Collector Pipeline for Metrics, Logs, and Traces
Build a high-performance OpenTelemetry Collector Contrib pipeline on Linux VPS, unifying OTLP ingestion, memory limiting, batching, and multi-backend export.
Observability8 min read
A One-Weekend Observability Stack That Stays Small
Assemble a weekend metrics, logs, and alerts stack on one VPS with hard RAM and disk caps so the observers never crowd out the application you run.
Observability9 min read
Uptime Checks From Outside the VPS Machine
On-box probes cannot see a dead host or a blocked NIC. Add a cheap external HTTP check so someone else notices when your Lucknow VPS stops answering.
Observability8 min read
Loki and Promtail on One VPS: When It Is Enough
Run Loki with a filesystem store and Promtail against journald on one VPS, with retention that fits the disk, and know the day this setup stops being enough.
Observability9 min read
logrotate: Keep VPS Logs From Filling the Disk
Configure logrotate so leftover text logs cannot fill a VPS root disk, with size caps, copytruncate versus create, and a dry-run habit before you need it.
Observability8 min read
journald as the Source of Truth, Not Log Files
Treat the systemd journal as the canonical log store on a VPS: persistent disk, field queries, vacuum limits, and one pipeline into Promtail or nothing.
Observability8 min read
Blackbox Exporter HTTP and TLS Probes on a VPS
Probe your own HTTP endpoints and TLS certificates from blackbox_exporter so Prometheus notices a dead site or an expiring cert before users do.
Observability8 min read
Alertmanager: Page Yourself Without Alert Fatigue
Wire Prometheus to Alertmanager with grouping, inhibit rules, and a for duration so a single VPS pages you for real faults, not for every blip.
Observability9 min read