February 2025 Summaries

Checkly, a monitoring tool, is expanding its capabilities to include network monitoring with TCP checks. This feature allows users to monitor critical non-HTTP services directly in Checkly without adding extra tools to their stack. With TCP checks, users can verify connections, detect failures proactively, and validate server responses, making it easier to ensure application reliability across the entire infrastructure. The feature is designed to be integrated seamlessly into development and monitoring workflows, allowing users to define, version-control, and automate TCP monitors alongside other monitors, and can be used for various applications such as mail servers, custom TCP services, firewalls, and network rules. Checkly's TCP checks provide advanced features like data exchange and assertions, private locations, and flexible fail conditions, making it a powerful tool for monitoring TCP-based infrastructure. The feature is included in the existing API check quota and is billed at $2 per 10,000 runs.

The Learning & Visiting budget at Checkly is a $1,000 annual allowance designed to support professional development and personal growth. The budget has flexible guidelines, allowing team members to invest in what matters most to them, whether it's books, courses, conferences, or tools. In 2024, the biggest categories of spend were coaching and mentoring (24.99%), tools (15.39%), subscriptions (11.43%), visiting (9.03%), conferences (7.84%), and training & courses (10.67%). Coaching and mentorship saw a significant increase in 2024 compared to 2023, reflecting the 70:20:10 learning model. The team used their budget for various purposes, including improving public speaking skills, acquiring new tech tools, attending conferences, and pursuing certifications. The budget has been found to have a positive impact on personal growth, confidence, and well-being, making it an essential part of Checkly's culture.

Checkly is expanding its monitoring capabilities to include TCP-based checks, allowing developers to monitor network-level resources and services that don't give full HTTP responses. This move shifts the responsibility for network service availability leftward, requiring developers to test and monitor their networks as part of their application development workflow. Checkly's new TCP monitors offer features such as SSL verification, setup and cleanup scripts, deep configuration options, and integration with existing monitoring tools like PagerDuty. With this addition, Checkly provides a more comprehensive monitoring solution that empowers developers to catch problems early and keep their entire system running smoothly, making it easier for teams to stay ahead of issues and maintain reliable applications.

Checkly enables engineers to automate the monitoring of their production services using Playwright, an automation framework. The goal is to find the right cadence for site checks, balancing frequency with minimizing unnecessary noise and ensuring timely alerts. This involves understanding service level agreements (SLAs), defining a mean time to detect (MTTD) failure, setting retry logic, scheduling dynamically based on region, and balancing with comprehensive monitoring. By refining the approach through feedback loops, engineers can continuously improve their monitoring strategy, ultimately setting themselves up for success in maintaining uptime, meeting SLAs, and delivering a seamless experience for users.

Checkly is a monitoring system that sends alerts when it detects failures, but ensuring those alerts are delivered and noticed by the team can be challenging. To address this issue, Checkly offers strategies such as using group-level alert settings to enforce broader policies, utilizing multiple alert channels including email, Slack, SMS, and webhooks, verifying checks are working properly before sending notifications, and implementing soft assertions to reduce alert fatigue. By applying these best practices, teams can stay on top of problems and keep their systems running smoothly.

Checkly is a key part of a professional developer’s workflow, making it easy to know if your service is up or down, and measure performance. As we integrate with almost any development workflow, we also have Prometheus endpoints to let you use the popular Grafana stack to keep track of your site checks’ status. However, large enterprise users grew in usage, causing their check performance data to grow, resulting in occasional 429 status codes. To optimize this process, Checkly added compression at a surprising layer of its stack, specifically using the Brotli compression algorithm on Node.js. This was done after testing showed that Prometheus response metrics can be big and compressing them would significantly reduce storage space and improve performance. However, Heroku dynos struggled to run the compression, leading the team to find an optimal configuration by optimizing the compression parameters, finding a sweet spot between CPU usage and quality of compression. By doing so, they were able to deliver better performance to their users without increasing infrastructure costs.