May 2014 Summaries

The text discusses how to effectively use Datadog's Events Stream feature to manage large amounts of events in an infrastructure system. It explains that the Events Stream creates a timeline of relevant events, but can become unwieldy with many events per unit of time. To address this issue, users can leverage the search and filter functionality within the Events Stream to focus on specific events or services. The text also demonstrates how to save these searches for future use and customize the Events Stream view using filters. Additionally, it explains how to include filtered events in a ScreenBoard for visualization purposes.

The Datadog Events Stream is a powerful tool that helps teams focus on the most important events in their infrastructure. By creating a timeline of notable events, teams can visualize and collaborate on issues in real-time. The Events Stream's search and filter functionality allows users to narrow down the noise and focus on specific services or servers. Users can save their searches for future use and customize their view with event filters, making it easier to identify key events and address them promptly.

This post discusses how to optimize Memcached performance by calculating hit ratios and timing individual requests using DogStatsD. The integration dashboard provides general Memcached performance metrics, but custom metrics can be reported using the client-side instrumentation library. By calculating hit ratios for specific requests, developers can evaluate cache effectiveness and ensure that Memcached is beneficial. Timing requests helps verify that Memcached is working as intended by comparing response times when the cache hits or misses. Overall, this approach provides valuable insights into Memcached performance within a web application.

The text discusses how to use Datadog and DogStatsD to optimize the performance of Memcached in a web application. It covers calculating hit ratios for individual requests, which can be used to evaluate the effectiveness of a cache, and timing requests to verify that Memcached is doing its job. By using these metrics, developers can gain insight into the performance of their cache and make data-driven decisions to optimize it further. The text also highlights the importance of considering other factors beyond hit ratios, such as request overhead and cache serialization costs, when evaluating a cache's effectiveness.

The integration of Datadog and Splunk allows users to view log alerts and reports created by Splunk in the context of infrastructure metrics and events captured by Datadog, enhancing visibility into infrastructure performance. This collaboration facilitates team communication about Splunk events, customization of event notifications, and correlation of Splunk events with infrastructure metrics for efficient root cause analysis. By using both tools together, users can identify the root cause of problems faster and more accurately.

Rudy Bunel introduces the new Datadog and Splunk integration, allowing users to view log alerts and reports in context with infrastructure metrics and events. The integration facilitates team collaboration over Splunk events and provides visualization of Splunk events within performance metrics dashboards. This enables more efficient root cause analysis by correlating Splunk events with other infrastructure metrics. The tool streamlines communication and analysis tools, such as PagerDuty and OpsGenie, to identify the root cause of problems faster and with greater accuracy. By utilizing this integration, users can gain better visibility into their infrastructure and improve overall performance monitoring.

Datadog is a popular metric collecting service used by Peloton Cycle to collect server and application metrics. The company uses Datadog's integrations with various software, including NGINX, to monitor system health and business indicators. However, they found that the out-of-the-box integration with Redis was not accurate enough for their needs, so they developed a custom solution using DogStatsD to monitor riders with structured data. Additionally, Peloton Cycle uses Datadog's AgentCheck framework to implement custom checks for replica or standby monitoring of their Redis and PostgreSQL databases. They found that Datadog is developer-friendly, easy to customize, and suitable for conveying system and business metrics to a company-wide audience.

The article discusses how to visualize StatsD metrics in two different ways - as counts or rates, using the example of queries per second over a couple of days. It explains that both graphs represent counts but in different and complementary ways. Counts allow quick viewing of the number of events in a short time span while rates suggest continuous change over time. The article introduces new tools to help choose the right visualization for StatsD rates and counters, showing how to calculate rates and counts with StatsD and how to create your own counts graphs using Datadog's Graph Editor. It also explains how to use these features in more advanced queries and how to visualize counts on screenboards. The article concludes by stating that the feature is currently limited to a subset of StatsD metrics but they are working on expanding its scope.

Datadog has introduced a new bar graph visualization for viewing metrics as counts, which allows users to quickly see the number of relevant events that occurred over a specified time interval. This is particularly useful for visualizing data with gaps between occurrences of events. In contrast, viewing metrics as rates suggests a value that is continuously changing over time and defaults to a stacked area visualization. Datadog's query engine aggregates data on a particular resolution, which enables users to choose the right visualization for their StatsD rates and counters. The Graph Editor has been updated to allow users to create custom counts visualizations by selecting the metric type, scope, groupings, and other settings. Users can also use this feature in the query JSON editor by adding `.as_count()` or `.as_rate()` to the end of the scoped metric definition. Finally, Datadog screenboards now support visualizing counts on screens using both timeseries and query value widgets.

OpsGenie's integration with Datadog enables the automatic syncing of alerts between both platforms, allowing OpsGenie customers to leverage Datadog's alert notification and incident response capabilities. This integration also facilitates the creation of new alerts in OpsGenie from within the Datadog UI, as well as the posting of OpsGenie alerts into Datadog for further discussion and analysis. The integration supports various communication channels and collaboration tools, including Slack, HipChat, and chatOps capabilities, allowing responders to assess situations, resolve issues quickly, and take preventive steps. By consolidating related alerts into one incident, teams can save time and ensure that no alert is lost, while also correlating metrics and alerts for easy monitoring of entire IT infrastructure.

SolveBio uses AWS Elastic Beanstalk to manage some of their apps on EC2, and they recently started using Datadog for monitoring and metrics. They faced issues integrating Datadog with Elastic Beanstalk due to separate installations of supervisord. To resolve this issue, they instructed Elastic Beanstalk to stop the Datadog Agent before deployment and start it again when done. The integration was successful by using environment configuration files provided by Datadog. Monitoring AWS Elastic Beanstalk with Datadog has been beneficial for SolveBio in gaining visibility into their app performance.

Elastic Beanstalk and Datadog are used by SolveBio to manage and monitor their applications on EC2, with a focus on Python and Node.js stacks. The company initially encountered issues integrating the two tools, but was able to resolve them by telling Elastic Beanstalk to stop and start the Datadog Agent during deployment. Once set up, the Datadog Agent can be deployed using environment configuration files and provides valuable insights into application performance, including metrics on AWS Elastic Beanstalk instances.

Memcached is an open source, high-performance distributed memory object caching system that acts as an in-memory key-value store for arbitrary data. It's commonly used to speed up dynamic web applications by caching the result of expensive queries made to a relational database, providing a significant boost in response time. Memcached stores data in memory, making writing and reading from the store extremely fast compared to disk-based operations, but the data is volatile and should not be used as a permanent data store. The Datadog Agent can collect key metrics from Memcached, including system memory, fill percent, evictions, hits, and misses, which are essential for monitoring the general health and usage of the cache. By installing the Datadog Memcached integration, developers can monitor these metrics and ensure that their cache is functioning correctly, helping to speed up their web applications.

Conor Branagan and Kevin Abraham discuss the benefits of integrating Fastly with Datadog, an edge cloud platform that includes a content delivery network (CDN), as well as services for image optimization, video streaming, cloud security, and load balancing. The integration allows users to visualize, analyze, and alert on Fastly metrics and logs in context with monitoring data from across their stack, enabling them to resolve issues before they degrade the user experience. With Datadog's tag-based alerting system, users can create sophisticated monitors for their Fastly services, which can notify them when specific metrics pass a user-defined threshold or behave anomalously. The integration also enables users to pivot seamlessly between Fastly metrics and related logs, allowing them to better understand concerning activity and troubleshoot effectively. Additionally, Datadog's Observability Pipelines allow users to extract key metrics from their Fastly logs without ingesting them into the platform, providing a cost-effective solution for managing log volume. Overall, the integration provides full visibility into the health and performance of Fastly services, enabling users to isolate bottlenecks, streamline their troubleshooting process, and reduce their mean time to resolution (MTTR).