August 2018 Summaries

In response to significant growth and the need for enhanced scalability, Quby, a developer of smart thermostats, transitioned from a traditional datacenter infrastructure to a cloud-based system on Amazon Web Services (AWS), using Docker and Mesos Marathon for orchestration. This shift necessitated a new approach to monitoring and securing their dynamic services, which they achieved by adopting Sysdig solutions. Sysdig Monitor and Sysdig Secure provided Quby with an efficient, cost-effective way to oversee their environment by offering a single instrument for monitoring, security, troubleshooting, and forensics. The integration of Sysdig reduced monitoring administration efforts by 400% and enabled Quby to quickly resolve issues while maintaining high transparency and confidence for their clients. This transition marked a move towards a DevOps-based organization, improving scalability and stability, and allowing Quby to demonstrate their infrastructure's reliability to potential customers.

In the second part of the guide on Kubernetes monitoring with Prometheus, the focus is on integrating additional components like AlertManager, Grafana, and PushGateway to create a comprehensive monitoring stack. AlertManager handles alert classification, grouping, and notification routing based on configured rules, while Grafana provides data visualization through dashboards, which can be auto-configured for convenience. The PushGateway is introduced for collecting metrics from ephemeral and batch jobs that can't be continuously scraped, and it ensures these metrics are retained for later retrieval. The guide also emphasizes the importance of configuring persistent storage for Prometheus metrics using Kubernetes PersistentVolumes, ensuring data retention across pod restarts. This part builds on the basics covered in the first installment, moving towards a more robust monitoring solution and setting the stage for automated deployment using the prometheus-operator in future discussions.