September 2015 Summaries

Sysdig has been recognized by Enterprise Management Associates (EMA) as a "Vendor to Watch" due to its innovative approach to container visibility, which aligns with the growing enterprise adoption of microservice-based architectures. As containers and microservices become more prevalent, there is a pressing need for monitoring and management tools specifically designed for these technologies, allowing organizations to deploy them confidently in production environments. Sysdig's architecture supports this demand by providing container-native monitoring solutions that offer insights into the containerized environments, complementing the shift from traditional physical and virtual machines to containerized deployments. This recognition underscores Sysdig's commitment to providing tools that enhance the deployment, maintenance, and scaling of modern applications.

Sysdig emphasizes the importance of monitoring key metrics for effective management of Cassandra clusters, drawing from their own experience using Cassandra in Sysdig Cloud infrastructure. Key metrics include read/write requests and latency, disk usage, CPU percentage, network bytes activity, file bytes, heap usage, garbage collection times, hinted handoff pending, and compactions. Sysdig Cloud offers a comprehensive monitoring solution by integrating these metrics into a single dashboard for easy analysis, enabling users to quickly identify performance issues by segmenting data by cluster or node. An example is given where Sysdig Cloud's dashboard helped resolve a compaction backlog issue by identifying a single node's configuration problem, leading to a quick resolution and the setup of an alert to prevent recurrence. This illustrates Sysdig Cloud's ability to provide granular visibility and streamline troubleshooting through automatic correlation with application and infrastructure metrics, even in containerized environments.

Sysdig Cloud has announced an expansion of its integration with Docker by adding comprehensive support for Docker labels, enhancing its container-native monitoring solution. Docker labels, which are metadata tags for images, containers, or daemons, allow users to organize and manage containers efficiently. The integration means that Sysdig Cloud can automatically incorporate Docker labels into its monitoring experience without requiring additional configuration, providing users with extensive context for monitoring and troubleshooting. This context includes correlating hundreds of metrics with metadata such as host names, IP addresses, AWS tags, and now Docker labels, offering unique visibility into container environments. This enhancement allows users to chart metrics like SQL response times or Java heap usage in real time and set alerts based on specific Docker label combinations, positioning Sysdig Cloud as a leading monitoring solution for container technologies.

Sysdig has released a significant update to its Sysdig Cloud user interface, enhancing usability based on user feedback. The update focuses on improving the Explore section, offering a larger workspace for infrastructure analysis and easy navigation through logically grouped views. The new interface also simplifies dashboard management with clear editing options and the ability to maximize screen space for better visibility. Additionally, the alerting system now includes State and Resolved statuses for notifications, allowing users to quickly identify and address performance issues with intuitive filtering capabilities. These improvements aim to provide a more efficient and user-friendly experience for Sysdig Cloud users.

In a discussion with Trevor Parsons, Co-Founder of Logentries, and Chris Crane, COO of Sysdig, the evolving use of containers for production workloads was highlighted, particularly their dynamic and ephemeral nature, which introduces unique challenges in networking, storage, security, and monitoring. As containers can scale rapidly and exist briefly, effective monitoring has become a critical concern, although recent advancements in APIs, new monitoring methods, and logging support have made it easier. The importance of collecting various metrics, such as application-level data, container stats, Docker API events, and system resource usage, was emphasized to gain a comprehensive understanding of system performance. The upcoming webinar promises to demonstrate how Sysdig and Logentries address these monitoring challenges by showing how to collect and correlate metrics across different levels of the container stack.

Sysdig has introduced new features, specifically the httplog and httptop chisels, to enhance its capabilities in decoding HTTP traffic on servers, marking a significant improvement over the previous echo_fds chisel. These tools allow users to gain insights into HTTP requests and responses, such as URLs and response times, providing a server-agnostic approach capable of inspecting inside containers. Httplog functions akin to a log by detailing each request with associated data, while httptop provides a summary view similar to the Unix utility 'top' for monitoring system processes. Sysdig's powerful filtering engine can be employed to focus on specific aspects, such as requests from certain containers or clients, making it a versatile tool for monitoring containerized web servers. These enhancements are part of Sysdig's ongoing efforts to expand its application-aware features, with the option to try Sysdig Cloud for comprehensive infrastructure monitoring.

Sysdig Cloud's newly introduced Overview page offers users a comprehensive, dynamically generated snapshot of their environment, providing easy access to high-level Key Performance Indicators (KPIs) and granular metrics without requiring additional setup or configuration. This feature aims to reduce what Sysdig humorously terms "mean-time-to-pretty-charts" (MTTPC) to zero, instantly displaying visually appealing charts upon installation. Users can explore various metrics, such as CPU, Memory, I/O, and Network Bytes, across containers and AWS services, facilitating a detailed understanding of performance hot spots and system activity. The platform's container-native approach ensures that all metrics, including those from within containers, are automatically correlated and easily accessible. The Overview page also includes a Latest Notifications section, allowing users to quickly assess and troubleshoot recent alerts. By integrating container and AWS CloudWatch metrics, Sysdig Cloud provides a holistic, real-time overview of an environment's health, enabling efficient monitoring and troubleshooting.

Falco Feeds enhances the capabilities of the open-source security tool Falco by providing continuously updated expert-written rules to address emerging threats. The latest sysdig release introduces expanded support for minimal container-focused Linux distributions, including RHEL Atomic Host, CentOS Atomic Host, and Fedora Atomic Host, alongside the existing CoreOS support. These Atomic Hosts are lightweight versions of traditional distributions designed primarily for running Docker containers, which traditionally lack comprehensive monitoring tools. By running sysdig within an Atomic Host, users gain full visibility into both the host and their containers. Sysdig, comprised of the sysdig-probe kernel driver and userspace programs, can now operate on distributions without available kernel headers by utilizing precompiled kernel modules, simplifying installation and enhancing troubleshooting capabilities in container environments. As sysdig continues to evolve, these improvements aim to advance the container ecosystem and facilitate easier monitoring processes.