MyDecisive today announced it has made its Smart Telemetry Hub available under a permissive open source license to drive further adoption of a platform based on Kubernetes that uses OpenTelemetry to collect telemetry data.
In addition, the company also revealed it has joined the Cloud Native Computing Foundation (CNCF) to influence the development of OpenTelemetry.
MyDecisive is making a case for an approach to collecting telemetry data that uses memory on the platform used to collect OpenTelemetry data to filter it, and any associated traces that need to be tracked, before being passed on to its observability platform. The overall goal is to reduce the volume of telemetry data that DevOps teams need to pass through to their observability platform.
MyDecisive CEO Ari Zilka said the company is joining the CNCF to press its case of expanding the existing OpenTelemetry standard to include instances that use local memory to filter data before passing it on to an observability platform. Additionally, DevOps teams can create intelligent alerts that monitor golden signals to track the health of specific services.
That approach also makes it easier for DevOps teams to ascertain the health of specific services in a way that serves to proactively reduce mean time to remediation, noted Zilka.
It’s not clear how much support there is in the OpenTelemetry approach being advocated for by MyDecisive, but many organizations are already struggling to manage the amount of telemetry data they are collecting. The issue is that OpenTelemetry was originally designed to simply collect data using open source agents that simply passed it through to a backend observability platform for analysis. It’s clear now there is a need to analyze that data earlier to reduce the volume of telemetry data that has become overwhelming as more applications are instrumented, said Zilka.
MyDecisive has not yet made any decision about contributing Smart Telemetry Hub to a consortium but it’s apparent that as the cost of processing and analyzing telemetry data declines, the number of applications that organizations will be willing to instrument is only going to increase. Historically, many organizations were reluctant to instrument every application simply because the cost of acquiring and deploying agent software to collect telemetry data was prohibitive.
It can still be challenging to deploy and manage OpenTelemetry collectors, but the real challenge now is reducing the total cost of observability by streamlining the management of telemetry data.
Each DevOps team will, of course, need to determine what percentage of their application portfolio needs to be instrumented. The one certain thing is that as application environments become more complex, the need to move beyond monitoring a set of pre-defined metrics is becoming more pressing. Hopefully, there will come a day soon when artificial intelligence (AI) makes observability more accessible to a wider range of DevOps teams. In the meantime, however, the more immediate challenge is aggregating all the relevant telemetry data in the most efficient and least costly way possible.