Thousands of customers love & trust Datadog
Most organizations operate multiple disconnected monitoring systems used by different teams to collect data from disparate sources in order to understand how an application is performing. Yet the various components of an application, their underlying infrastructure, and the teams that operate them work together in interconnected workflows to complete processes from beginning to end. Thus, for teams to effectively work together as they iterate on an application, aim to fully understand its health, and troubleshoot any issues that arise, they require access to the performance information from all relevant parts of the application. Without a single platform to unify the performance data that is collected in multiple places, this becomes challenging — if not impossible.
Datadog provides a single source of truth of performance data for an entire organization — unifying logs, application traces, infrastructure metrics, and browser and API synthetics tests. Performance data is universally available to all users in the same UI, and all performance data is interconnected and can be compared, correlated, and combined for analysis, visualization, and alerting. Ultimately, Datadog provides universal context and data access so that any user can self-serve the answers to their questions about performance and health.
“The single most exciting part about Datadog is that we can have a single pane of glass, with everything that we need as far as Serverless, logs, metrics, third-party integrations, that all come together and we can quickly find out if there's any issues in our stack. It's phenomenal — I don't need to go into two or three different applications to see what's going on here or there, and try to correlate, everything is in Datadog, right there.”
– Reza Javidi, Director of DevOps and SRE, Openfit
Engineering organizations have traditionally relied on four (or more) separate systems to monitor their front-end websites, application code, logs, and underlying IT infrastructure. All four of these data sources provide crucial insights into the health of different parts of an application stack that are interconnected, making visibility into each mission critical to ensure an optimal customer experience. However, because monitoring solutions for each data source come in best-of-breed systems focusing only on one data source, understanding how an application is functioning from front end to infrastructure has previously required “eyeball” analytics from several screens patched together, integrations between unlike products that miss crucial interconnections, or external analysis with complex data aggregation tools. The end result is that issues which span multiple parts of the stack are time-consuming and tedious to troubleshoot, or are never fully understood at all. Datadog provides a single platform with full data and UI interconnectivity that bridges together logs, synthetic tests, application traces, and infrastructure metrics. The end result is a unified view of the entire stack where issues can be troubleshot instantaneously in one system even when multiple components — such as a field on the front-end of an application, a subroutine in the application’s business logic, and a single host in the back end — are affected.
Whether a company has allowed teams free rein over their tooling, or new divisions have been brought in via acquisition, many organizations find themselves with a patchwork of disparate monitoring tools spread over departments. This multiplicity of solutions, however, slows down engineering initiatives and issue resolution when teams need to collaborate. Because each team has its own source of truth, simply understanding the health of an application requires negotiation and can devolve into finger-pointing. When the many monitoring tools in use don’t fit together, making sense of what is occurring in joint workflows becomes a project within itself, and may not even be possible if certain data is not collected by a downstream tool. Further, fully onboarding team members typically requires training in multiple tools. This doesn’t usually happen, and new team engineers are often left blind to the performance of parts of their application.
Datadog allows for multiple teams to work off the same monitoring platform that each team can then tailor to its needs. With more than 350 vendor-supported integrations, every team that onboards onto Datadog can integrate the data from their systems into a unified platform within minutes. Data from all teams is then universally available to all users, so that teams can draw in data from upstream or downstream systems and “see around the corner,” as it were. Dashboards and alerts are highly configurable and shareable, and individual users can lock their objects for editing. Datadog also has parent-child federation capabilities, so that sub accounts can be deployed for individual divisions or departments by IT or another central department. The parent account can - oversee the child accounts and draw in anything that is being collected by each child account, allowing for federated alerting and dashboarding that is critical to centralized IT and support departments.
For organizations using a multitude of monitoring tools, it is unlikely that any team member will be able to have access to all of the data to understand the performance impacts of their work. In such situations, the specialized work from a small number of administrators of the monitoring is required to gain this visibility. Because of this extra work, and typically, wait times for reports, it becomes challenging to create a culture of performance accountability within an engineering team — out of sight, out of mind. Datadog was designed to be universally used within an organization. Users onboard themselves in minutes and have instant access to all dashboards and alerts used by their teams, - as well as the data for all of the systems that their teams and - other related teams employ. With Datadog’s federation capabilities, an organization can have multiple Datadog child instances to ensure that the data access between teams is executed correctly. The end result is a culture of performance accountability easily built with Datadog, as every piece of work that an engineer undertakes can be immediately set up for monitoring by the engineer that is doing the work.
Applications typically produce a number of unique, proprietary metrics, such as items in a cart, successful credit card transactions, check-ins, or ads served, related to the application’s business functions. This data is critical to understanding the business performance that went through an application, and can act as an early warning of issues if the data patterns begin to deviate. Engineering teams often homegrow tooling to collect this data, and find that the maintenance to keep up with the growth in the application either becomes prohibitive or requires an ever-expanding support team to continue sending this data on. Datadog includes robust collection mechanisms for custom business metrics from applications. This data is immediately available in Datadog to analyze, visualize, and compare with application and infrastructure data when troubleshooting issues. Importantly, this data can be alerted on, to provide immediate notification of business-impacting conditions, such as a drop in shopping cart checkouts or a suspicious spike in ad clicks. Ultimately, organizations are able to use Datadog organization-wide to collect these business metrics and retire these “DIY” tools.
Datadog provides a robust set of standards to collect data. In addition to more than 350 vendor-supported integrations to collect infrastructure metrics, logs, discrete events, and application traces as applicable per integration, Datadog can also take in data via API, SNMP, WMI, JMX, StatsD, FluentD, and CollectD. This allows for nearly any application-related data to be ingested. Once this data is ingested, it is available for analysis with Datadog’s extensive graphing widgets or can be set up for notification when certain conditions arise with Datadog’s alert triggers. These capabilities allow for application data to be processed in creative ways that may plug gaps in workflows or use cases which previously had no visibility or connectivity between systems. Ultimately, Datadog may be able to provide solutions in areas that would not be thought of as traditional IT “monitoring.”
Datadog also provides a great deal of functionality to get insights and data to external stakeholders. From publicly available read-only dashboards to customizable notification paths for alerts and a Webhooks integration that can be triggered to undertake predefined, automated actions, Datadog can flexibly share data with access controls defined by the organization.