Lightning Talk: Locating and Debugging Failures with Linkerd and Telepresence

Linkerd and Telepresence offer a simple way to better observe and debug applications running in Kubernetes clusters.

• Cloud native applications can become hard to maintain and evolve
• Linkerd and Telepresence provide a solution for observing and debugging applications
• Sample code is available for easy replication
• An anecdote was given to illustrate how Linkerd and Telepresence can be used to locate and fix problems in a cluster

The presentation gave an example of an app that allowed users to vote for emojis, but users were complaining about problems in the app. Linkerd was used to identify the location of the problem in the cluster, and Telepresence was used to debug and fix the problem in the voting service. A special http header was used to ensure that testing and debugging did not affect other developers on the team.

Cloud native applications that consist of many services running in K8s clusters can become hard to maintain and evolve. From our experience and what we see in the Kubernetes community Slack channels, when developers hit the tipping point of not being able to run all their services locally, they run into trouble. At this time, they suddenly realize they need new tooling and approaches for observing and debugging applications that will spread across their local dev machines and remote clusters. This presentation will show operators and developers how Linkerd and Telepresence offer a simple way to better observe and debug applications running in their clusters. Easy to follow and replicate (with sample code made available), users will be able to get results quickly, without disrupting their existing services architecture.Click here to view captioning/translation in the MeetingPlay platform!