June 2017 Summaries

Nuno Sousa, Senior Developer at Uphold and creator of Datadog's integration with Home Assistant, discusses how Datadog can enhance the capabilities of Home Assistant. Home Assistant is an open-source platform for tracking and automating home devices, while Datadog allows users to aggregate, analyze, and monitor data from Home Assistant. By sending Home Assistant metrics to Datadog, users can detect anomalies in temperatures, correlate events with device malfunctions, set alerts for Z-Wave message relay failures, and identify trends in power usage. The integration involves setting up the Datadog component, which listens for state changes in devices/entities and pushes their latest attribute values to the Datadog Agent. Users can then create custom dashboards to visualize data from multiple Home Assistant components, such as temperature, humidity, battery levels, power metrics, and more. To configure the integration, users need to install the Datadog Agent on a host and add specific configuration details to their Home Assistant configuration.yaml file.

Nuno Sousa, a Senior Developer at Uphold, explains how Datadog integrates with Home Assistant, an open source platform for tracking and automating home devices. With this integration, users can aggregate, analyze, and monitor data from their IoT devices in Home Assistant, enabling features such as anomaly detection, correlation of events with device malfunctions, and setting alerts for device issues. The integration allows logbook entries to be sent as events to Datadog, allowing users to correlate between triggered automations and metrics. Users can also visualize their data in dashboards, creating customized graphs and charts to display metrics across multiple Home Assistant components. To configure the integration, users need to add a specific configuration to their Home Assistant `configuration.yaml` file, specifying the IP address of a host running the Datadog Agent or using default values. Once configured, the integration enables users to monitor and analyze data from their IoT devices in Home Assistant with ease.

Datadog recently held an internal hackathon where teams were tasked with creating projects of their choice for two days. One team decided to build a way to display Datadog graphs and dashboards inside the Minecraft game world, allowing engineers to monitor applications and infrastructure while gaming. The project involved controlling Minecraft via Python using Raspberry Juice and PI Edition, accessing real-time data from Datadog, and integrating the two pieces. Challenges included configuration management, persistence, and displaying data efficiently. Overall, the hackathon was a fun and enjoyable experience for participants.

Datadog recently held an internal hackathon where two teams, led by Christian Mauduit and Corentin Dancette, decided to build a way to display Datadog graphs and dashboards inside the world of Minecraft. Their goal was to programmatically access Minecraft elements, retrieve real-time data from Datadog, and glue it all together to read data from Datadog and pump it into Minecraft. They used Python as their primary language due to its familiarity and existing libraries. The team faced several challenges including configuration, persistence, and displaying data in a viewable manner, but ultimately succeeded in creating a functional project that displayed metrics in real-time with monitors that would shine red when in alert mode. The hackathon was a fun experience for the team, allowing them to play around with new technologies and enjoy the process of hacking for pure fun.

Cloud Foundry is an open-source platform that enables the development and deployment of applications on any infrastructure while supporting distributed systems and microservice architectures. Datadog has introduced a new integration for monitoring Cloud Foundry clusters, which helps maintain the health and performance of applications. The integration collects metrics from critical Cloud Foundry components such as BOSH, Cloud Controller, Loggregator, router, and Diego to visualize their performance, identify patterns and anomalies, and create alerts. To start monitoring Cloud Foundry clusters automatically, users can upload the Datadog Agent release to their BOSH Director, configure it as an add-on, and add a UAA client for the Datadog nozzle.

Datadog has announced a new integration with Cloud Foundry, designed to automatically monitor the health of Cloud Foundry clusters. This integration allows operators to collect metrics and logs from critical Cloud Foundry components, including BOSH, Cloud Controller, Loggregator, router, and Diego. The Datadog Agent collects these metrics and visualizes their performance, enabling users to identify patterns and anomalies, create alerts, and maintain the health and performance of their applications. To get started, users can upload the Datadog Agent release to their BOSH Director, configure it as an add-on, and add a UAA client for the Datadog nozzle. If already a Datadog customer, they can start monitoring Cloud Foundry clusters immediately; otherwise, signing up for a free trial is recommended.

Datadog introduces a new Manage Monitors UI to simplify the process of finding, filtering, and organizing monitors for organizations of all sizes. The updated interface includes advanced search capabilities, enabling users to locate specific monitors by name or attribute. Additionally, multi-edit functionality allows for efficient bulk operations on multiple monitors at once. This new feature aims to improve incident response times and provide better visibility into monitoring coverage across various services and teams.

The new Manage Monitors UI in Datadog aims to make it easier for teams and organizations of all sizes to find, filter, and organize their monitors. A new sidebar allows users to quickly filter on different dimensions and attributes of their monitors, making incident response faster and more efficient. Advanced monitor search has also been introduced, enabling users to locate monitors by name or attribute in a faster and more precise manner. The updated page is also significantly faster, with improvements that will benefit organizations with large numbers of monitors the most. Additionally, multi-edit functionality has been added, allowing users to tag multiple monitors at once and apply bulk operations such as muting and resolving, further enhancing the user experience.

Recently, the Datadog Agent was extended to support extracting additional metrics from Kubernetes using the kube-state-metrics service. Metrics are exported through an HTTP API that supports content negotiation so that one can choose between having the response body in plain text format or as a binary stream encoded using Protocol buffers. The article provides a detailed introduction to Protocol buffers, including its advantages and disadvantages. It also discusses benchmarking results for encoding and decoding messages with Protobuf and plain text formats. In conclusion, the use of Protobuf can significantly improve performance when dealing with complex data structures in Kubernetes state checks.

The Datadog Agent has been extended to support extracting additional metrics from Kubernetes using the kube-state-metrics service. This is achieved through an HTTP API that supports content negotiation, allowing for plain text or binary format output. Protocol Buffers (Protobuf) are used for binary encoding and decoding, offering efficient data transfer. However, due to Python implementation limitations, a C++ extension is recommended for performance gains. Benchmarks show Protobuf outperforming plain text in terms of payload size but the pure Python version is slow. Despite this, implementing chained messages with Protobuf requires minimal code changes. The new `kubernetes_state` check released with Agent version 5.10 uses Protobuf for efficient data processing and decoding.