Overview
Container Observability is an intelligent observability tool for modern cloud-native applications built from the ground up for Kubernetes platforms and applications. The platform uses a broad set of telemetry, as well as configuration and change information to build an understanding of the application. Using contextual ML, Container Observability can predict performance degradations in real-time before the users notice. Further, providing automated causal analysis and fault isolation capability can improve Ops staff productivity.
With Container Observability, there is no need to deploy proprietary agents or code instrumentation.
Besides future-proofing your observability platform, you gain several advantages, including:
Applications can be on the public cloud, on-premises, or a hybrid of both and still be observable with a single platform.
Using OpenTelemetry (OTel) and open-source tools means there is no need for proprietary host agents.
Data collection is lightweight as only a single pod collector of Telemetry by type per cluster is all you need.
Monitored data stays in your control as the application owner and can be used for multiple business purposes.
Cost savings are significant as the stored telemetry can utilize low-cost cloud storage (e.g., AWS S3) as opposed to a vendors’ monitoring services.
Accessing telemetry does not incur cloud ingress charges.
Pick from the many deployment options including SaaS, on-prem software.
With access to all monitoring data from open-source instrumentation, Container Observability unifies, i.e., contextually “stitches together” all telemetry to extract and provide deep insights.
With Container Observability you can:
Auto-discover all full-stack dependencies - between services (‘service map’) and from services down to orchestrator and infrastructure (‘three-layer view’).
Provide a single UI into different facets of the application from telemetry, performance, configuration, and changes.
Expose dependencies in the application in real-time through Flow and Trace Analytics.
Deliver predictive anomaly detection without requiring manual Ops intervention.
Automate causal analysis to isolate problem components.
The above capabilities are provided continuously in real time by using a multi-stage pipelined process. To improve the efficacy of analysis in each stage, Container Observability embeds and uses curated knowledge on the technology stack, known applications, and IT diagnostics processes.
Related topics: