January 2017 Summaries

The text discusses the importance of monitoring Ceph, a storage infrastructure used in Kubernetes and OpenShift, highlighting the necessity of tracking its performance and health. It delves into using both built-in command-line tools and Sysdig Monitor for effective monitoring. Built-in tools provide basic status checks such as health, OSD node status, and capacity usage, while Sysdig Monitor offers advanced features like historical metrics, auto-discovery, and metadata tagging from various platforms. The article recommends setting up alerts for key metrics, such as health status, monitor node quorum, OSD node participation, storage capacity, and time synchronization to ensure efficient Ceph cluster management. It concludes by suggesting that while built-in tools suffice for smaller, non-critical applications, a more robust system like Sysdig is beneficial for production environments to provide comprehensive monitoring and alerting capabilities.

Published by Jorge Salamero Sanz, the blog post focuses on the use of Ceph as a persistent storage solution in Kubernetes or OpenShift environments, emphasizing its flexibility, performance, and advanced features like snapshots. The author details the process of setting up a Ceph cluster and configuring Kubernetes or OpenShift to store volumes as block storage images, highlighting steps such as creating a dedicated pool, setting up block device images, and configuring client packages on cluster nodes. The blog post provides a comprehensive guide on creating and managing Kubernetes entities like secrets, persistent volumes, and persistent volume claims, illustrated through YAML configurations. A practical example is demonstrated by deploying a MySQL Pod that utilizes the Ceph storage, with instructions on mounting the block device as the Pod transitions across the cluster. The article concludes by underscoring the importance of monitoring Ceph's health and performance before moving to a production environment, setting the stage for a subsequent discussion on monitoring strategies.

Ceph is a self-hosted distributed storage system that is increasingly popular among organizations using Kubernetes and OpenShift due to its scalability and flexibility. This guide provides a comprehensive overview of deploying Ceph on AWS, including strategies for using Ceph as a storage backend for Kubernetes or OpenShift Persistent Volumes. It highlights Ceph's object, block, and Posix-compatible network file system storage methods and explains the CRUSH algorithm used for data placement. The guide also discusses Ceph's basic terminology and alternatives, such as GlusterFS and NFS, while detailing how to deploy Ceph using tools like Ansible and ceph-ansible. Additionally, it addresses the challenges of containerized Ceph deployments and offers insights into monitoring its performance using tools like Sysdig Monitor.

Sysdig, an open-source tool, was developed to provide comprehensive visibility into Linux systems by bridging the gap between network activities and machine processes, a challenge not effectively addressed by existing tools. The updated Sysdig cheatsheet offers commands that enhance understanding and troubleshooting of containers and Linux servers, making it a beneficial resource for users transitioning from traditional tools like htop, strace, and tcpdump. Sysdig's creation coincided with the development of Docker by DotCloud, marking a pivotal moment in container technology. Additionally, Sysdig's Falco Feeds offers continuously updated, expert-written rules to help open-source-focused companies stay ahead of emerging threats.

The winter 2016-2017 release notes for Sysdig Monitor highlight several key updates and features aimed at improving container monitoring and user experience. Sysdig Teams was launched to facilitate the monitoring of containers and microservices using Kubernetes metadata to define visibility and access control, particularly benefiting users with complex infrastructure needs. New dashboard panel scopes allow for individual metric configuration, and the alerting interface was redesigned for simplicity, with default alerts for new accounts covering essential system metrics. Additionally, custom App checks can now be configured per container, enhancing monitoring flexibility. Improvements also include a more intuitive time navigation UI, new metrics, enhanced file system usage metrics, and a new rate aggregation option for evaluating metrics. Sysdig Monitor and open-source Sysdig are now available in the Rancher Catalog for easy deployment, and a new status page keeps users informed about system performance and events. Integration with OpsGenie for native alerting and an in-app invoicing feature for billing convenience are also introduced, emphasizing Sysdig’s commitment to user feedback and continuous improvement.

Falco 0.5.0 has been released, introducing a variety of updates including new security monitoring features, rule changes, and bug fixes aimed at enhancing its functionality as a behavioral security monitor. The release includes optimizations to reduce notification noise, such as caching output formatting objects and implementing a token bucket for notification throttling. New features allow Falco to record event processing statistics and run in a container without explicitly loading the kernel module, while log verbosity can be adjusted through configuration. The ruleset has been updated to include new programs and images, and bug fixes ensure better validation of output fields and the functionality of build options. The update is accessible through rpm/debian packages, Docker Hub, and GitHub, with further details available in the changelog.

The blog post by Kamol Mavlonov provides a detailed guide on setting up dashboards and alerts to monitor resource utilization in Mesos environments, specifically focusing on CPU, memory, and disk usage. It explains how to create a disk utilization dashboard by selecting various file system metrics and setting color-coded thresholds for usage. Additionally, the post outlines the process for creating alerts when specific usage thresholds are exceeded, such as 60% for disk utilization, 95% for memory utilization, and 90% for CPU utilization. The guide emphasizes selecting appropriate scopes, conditions, and metrics for each alert, enabling notification channels, and optionally using Sysdig capture for automatic monitoring.