How can version aware distributed trace analysis be automated?

How to automate version aware distributed trace analysis

From a handful to millions of distributed traces requires automation

When you capture distributed traces in shared testing or production environments with hundreds of services/microservices deployed in one or more versions, you could quickly wind up with millions of collected fractions. If you can automate the study of those traces, you can enable DevOps and SRE teams. They will be able to answer queries like:

• “Which versions of our services are handling our essential transactions right now?”
• “How does an overburdened backend service affect the frontend service’s SLOs?”
• “Which frontend services are to blame for the backend services’ changing traffic behavior?”
• “Do two different versions of a service behave differently?” If that’s the case, should we halt the production rollout?”

These are questions I’m hearing more and more these days, which is a little concerning. You need to record version information on every distributed trace to address version-specific questions like the one above. Still, it would help if you also automated the analysis because no one can manually sift through millions of trails and come up with answers that lead to better delivery and release decisions.

About Enteros

IT organizations routinely spend days and weeks troubleshooting production database performance issues across multitudes of critical business systems. Fast and reliable resolution of database performance problems by Enteros enables businesses to generate and save millions of direct revenue, minimize waste of employees’ productivity, reduce the number of licenses, servers, and cloud resources and maximize the productivity of the application, database, and IT operations teams.