Grafana 依靠 Embrace 来获取移动数据_AI阅读总结 — 包阅AI

Grafana Labs has opted to deepen its integration with Embraceas a way to enhance its capabilities in mobile application observability. Previously, Embrace was available as a Grafana plugin, enabling real-time data visualization of data stored in Embrace and analysis for mobile apps. However, this new integration offers much more.

With the integration of Embrace, users can gather complete frontend telemetry data from mobile apps. This enables the analysis of mobile app application performance data directly in Grafana Cloud by ingesting metrics, traces and (in the future) logs directly from Embrace. Its reach extends to the end user, logging and analyzing error messages on devices, whether iPhones, Android phones or others.

Going beyond its limitation as a plugin means Embrace and Grafana Cloud customers can now write queries “across all of their telemetry data, from infrastructure to the mobile application, and build a whole system view into their SLOs [service-level objectives] and user experiences,” Logan Smith, director of business development, told The New Stack.

Grafana customers had been asking for ways to instrument and observe their mobile applications to provide a complete view of their infrastructure and apps, and then measure and improve SLOs and user experiences, Smith said. Embrace builds its own distribution for OpenTelemetry (OTel), so teams have a variety of implementation paths, but when using it with Grafana Cloud they get pre-built experiences to help mobile engineers participate in an organization’s observability practice, Smith said.

“Grafana works for all mobile application observability. With Embrace, we provide a happy path of deeper integration and shorter time to value,” Smith said.

Embrace itself is not open source per se — it’s a closed-source platform that builds open source instrumentation. Its SDK used to instrument the app is open source software and based on OpenTelemetry, Smith explained. Mobile telemetry data has its own set of considerations — intermittent connectivity to the device, type of usage, etc. — and it needs to be handled differently, Smith said.

“This just underscores the importance of our ‘Big Tent’ philosophy and why we prioritize interoperability within the wider ecosystem and with tools like Embrace,” Smith said. “We believe that organizations should own their own observability strategy, choose their own tools and have the freedom to bring all their data together in one unified view in Grafana.”

With Embrace, data now pushes to your Grafana panel for debugging. This capacity offers both accessibility and visibility for individual actions, error messages, timing, timeouts and other aspects of applications in real time. For example, it is possible to home in on incidents as they occur and correlate possible connections between them. Each incident can be pinpointed and analyzed individually, along with the entire usage timeline of the application.

In the event of a crash, it can be quickly determined whether particular errors are widespread or limited to specific use cases in real time. By using the capacity to pinpoint and analyze specific errors, it’s possible to focus on particular crashes or errors and see how they might correlate with certain patterns, aiding in the analysis of larger problems.

The OTel Connection

This capability is built into the application, with data pushed through OpenTelemetry via the Grafana panel. A key challenge the Embrace and Grafana Cloud offering helps to resolve with OpenTelemetry is the gap between mobile application and backend infrastructure observability data, Smith noted, which generally makes it harder and more time consuming to pinpoint and resolve issues.

According to the Grafana 2024 Observability Survey, 85% of respondents are investing in OpenTelemetry, and its popularity extends to both application and infrastructure observability, he said.

“This means OTel, with its open standards and vendor neutrality, is the clear choice for bridging that gap,” Smith noted. “Because Embrace’s SDKs rely on OpenTelemetry to ship telemetry data to Grafana Cloud and Grafana makes interoperability with OTel easy, this combination provides a powerful observability solution without fear of vendor lock-in.”

Why It’s Important

Mobile observability is certainly critical as mobile applications continue to play a key part in distributing computing and networking. To that end, Grafana, with Embrace in this case, has very good reasons to adapt its solutions for telemetry data pull along with other leading observability players including Splunk, Datadog, Honeycomb and others.

Real-user monitoring is a good example. As Gartner describes it, real-user monitoring measures the experience of application users. JavaScript is automatically injected into web applications to collect data, Gartner analysts Padraig Byrne, Matt Crossley, Martin Caren and Mrudula Bangera co-wrote in their report, “Key Considerations When Choosing APM and Observability Platforms.”

Alternatively, plugins can be deployed when HTML is not accessible (such as in the case of SaaS applications), they write.

“Session recording and replay shows what the user experienced during an application session, which the IT monitoring teams can correlate with other application performance metrics for root cause analysis (RCA),” the analysts write. They say the target metric communicates:

• Basic: JavaScript injection, collection of user data such as applicationresponse time, latency, errors, geolocation, browser and versions.
• Advanced: Support for mobile and non-browser-based applications, visualizations of the end-client journey, including analysis of business metricssuch as client spend.