Best of Monitoring — October 2024

  1. 1
    Article
    Avatar of javarevisitedJavarevisited·2y

    Optimizing Node.js Performance: Tips and Tricks

    Node.js is renowned for its speed and efficiency in handling multiple simultaneous connections. To ensure your Node.js applications are optimized for performance and scalability, it's crucial to understand and monitor performance metrics like response time, requests per second, CPU and memory usage, and error rates. Key techniques include avoiding blocking operations, minimizing memory usage, implementing caching strategies, and profiling and optimizing code. Furthermore, optimizing database queries, using connection pooling, and employing CDNs can significantly enhance performance. Understanding the event loop and employing best asynchronous programming practices are also critical. Lastly, proper server configuration and deployment practices can help maintain and improve application performance.

  2. 2
    Article
    Avatar of faunFaun·2y

    Learning Go by Instrumenting a Go Application for Prometheus Metrics

    A beginner's guide to learning Go by instrumenting a Go application for Prometheus metrics. This tutorial covers building a Prometheus metrics exporter to consolidate analytics and metrics from Datadog's SLO product. Key steps include parsing the Datadog API response, creating necessary structs in Go, declaring Prometheus metrics, initializing the Datadog client, fetching SLO data, and pushing the metrics to Prometheus.

  3. 3
    Article
    Avatar of grafanaGrafana Labs·2y

    How to use Prometheus to efficiently detect anomalies at scale

    Discover how an effective anomaly detection framework was built using Prometheus and PromQL. Learn about setting up average and standard deviation recording rules, tuning parameters like time windows and multipliers, and addressing challenges such as extreme outliers, low sensitivity, and seasonality. The reusable framework works for any metric and can integrate with your existing Prometheus setup to enhance incident investigation and root-cause analysis.

  4. 4
    Article
    Avatar of last9Last9·2y

    Prometheus Alertmanager: What You Need to Know

    Prometheus Alertmanager helps manage alerts in a production environment by organizing, routing, and deduplicating alerts, thereby reducing alert fatigue. It supports features like alert grouping, silencing, inhibition, and high availability setups. To effectively use Prometheus Alertmanager, ensure to configure alert conditions properly, use grouping and inhibition to avoid notification spam, and implement security best practices such as authentication and TLS encryption. Periodically review and audit alerts to keep configurations relevant and improve upon past incident learnings.

  5. 5
    Article
    Avatar of grafanaGrafana Labs·2y

    Monitoring Kubernetes: Why traditional techniques aren't enough

    Kubernetes offers significant advantages for large-scale deployment and management of applications, but traditional monitoring techniques fall short. Observability now leverages out-of-the-box solutions like Prometheus, Grafana, and OpenCost to facilitate proactive monitoring, cost management, and better resource allocation. The Kubernetes ecosystem makes it easier for teams to support application performance, though engineers must still be vigilant about costs and reliability.

  6. 6
    Article
    Avatar of faunFaun·2y

    Kubernetes Go-live checklist for your Microservices

    Successfully taking Kubernetes microservices from development to production requires careful planning and configuration. Key steps include calculating per-pod capacity, setting resource requests and limits, configuring autoscaling, ensuring high availability, implementing probes for health checks, and setting up comprehensive monitoring and alerting systems. By following this checklist, you can enhance the stability, scalability, and performance of your applications.

  7. 7
    Article
    Avatar of communityCommunity Picks·2y

    castai/egressd: Kubernetes aware network traffic monitoring

    castai/egressd is a Kubernetes-aware network traffic monitoring tool that uses a DaemonSet pod on each node to fetch conntrack entries for pods. It supports both Cilium eBPF maps and Linux Netfilter Conntrack module. The tool adds Kubernetes context to traffic records and can export logs to HTTP or Prometheus. Egressd operates as a privileged container to perform DNS tracing and conntrack entry fetching. The post includes a demo setup with Grafana and Prometheus, and additional instructions for exposing Grafana locally and running end-to-end tests.

  8. 8
    Article
    Avatar of grafanaGrafana Labs·2y

    Key Prometheus concepts every Grafana user should know

    Prometheus is a robust monitoring framework, far beyond just a time series database, offering features like metric scraping, powerful querying (PromQL), alerting, and service discovery. Key components include client libraries for instrumentation, the node exporter for OS-level metrics, and general exporters for diverse sources. PromQL enables complex queries and supports recording rules for optimizing data retrieval and storage. Prometheus excels in dynamic environments, adapting well with Grafana for enhanced data visualization. Understanding these capabilities, especially through resources like 'Prometheus Up & Running,' can significantly improve your observability practices.

See all Monitoring archives