Why Companies Waste 25% of Their Infrastructure Budget on Garbage Telemetry

本站添加: 2026-05-17

[00:00:00]Before I started the 30 days of Open Telemetry, I assumed it was a solved problem. You use the vendor or use use Open Telemetry. You plug it in and you're done. I have been doing observability both on the receiving and the giving end for almost 15 years now. And competition among observability vendors 10 years ago, but also 5 years ago, was very different. It was about who has the best data. And most of the investment in research and development was about collecting telemetry from all types of systems and stuff that was put in production when I was still in kindergarten. Open Telemetry has changed that. Open Telemetry, Prometheus, and other open source projects to the extent where now the competition is about of about who does the best with the data that everybody has. The fact that now there is a lot of high-quality open source technology to collect a lot of telemetry means that there is a lot of telemetry out there. And people are drowning in it. From a structural standpoint, the observability industry is predicated and its its growth and its continued growth to to please the the stakeholders, investors, and everybody else is largely predicated on generating more telemetry. There is little incentive out there for vendors to help you reduce telemetry. The only niche of the industry that uh does that is the observability pipeline tools where you pay money to them to pay less comparably less money to the observability vendor where the data goes. A lot of the feature functionality inside the seller is predicated on the idea that you should be able to get rid of garbage telemetry. Mhm. And statistically, most telemetry's garbage. You will never look at it. It's it's entropy.