February 2016 Summaries

To help users make sense of their systems quickly and eliminate ambiguity in metric interpretation, a comprehensive metadata catalog has been introduced for approximately 3,000 metrics collected by supported integrations. Metadata units are displayed automatically on timeseries graphs, query value widgets, and toplists, with descriptions available under the new "Metrics" tab for each integration in the Datadog app. This feature is now available for all standard integrations of existing Datadog customers, while those without an account can sign up for a full-featured 14-day trial.

To help developers quickly diagnose issues in production, Datadog has rolled out a comprehensive metadata catalog covering standard metrics collected by its supported integrations. This catalog provides measurement units and brief descriptions for approximately 3,000 metrics that surface automatically on graphs and dashboards. The metadata is displayed automatically on timeseries graphs, query value widgets, and toplists, with units shown in fractions of a second to milliseconds, millions of bytes per second to MiB/s, etc. Developers can find a complete list of collected metrics, their units, and descriptions under the "Metrics" tab for integrations in the Datadog app or in the docs as well. The feature is available for all standard integrations, including those for popular systems such as Redis.

Monitor View is a new feature offered by Datadog that provides a graphical representation of how data is processed by alerting backends, improving the ability to predict when alerts will trigger. This feature is particularly useful for scenarios where alerts are set based on moving averages, which can be challenging to compute intuitively. By using Monitor View, users can clearly see when specific metrics, such as the example given with the `haproxy.backend.queue.time`, exceed configured threshold values, offering a clearer understanding of alert conditions. Datadog customers can access this feature by navigating to the Monitors page and switching from Original Data to Monitor View, and new users can explore it by signing up for a free trial.

OpenStack uses terms like host aggregates, flavors, and availability zones in unique ways that can lead to confusion. Host aggregates are used by administrators to group hardware according to various properties such as RAM size, while availability zones are customer-facing and usually partitioned geographically. Flavors are the customer-accessible subset of host aggregates, defining specific public flavors from which clients can choose to run their virtual machines. Understanding these distinctions is crucial for both end users and administrators when using OpenStack services.

OpenStack's host aggregates are used by administrators to group hardware according to various properties, such as physical host configurations, and are not visible to customers. In contrast, availability zones are customer-facing and usually partitioned geographically, while flavors are public subsets of host aggregates that customers can choose from to run their virtual machines. Understanding the distinction between these terms is essential for planning OpenStack deployments and choosing the right configuration for specific use cases.