Do you remember that microservice that you wrote and successfully deployed in production a few weeks ago? I am afraid to tell you that it is under performing and occasionally generating faults that are compromising the rest of the system. Could you please join our war room in an hour to help us fix this as quickly as possible?
For most developers this reads like the plot of a horror movie. There is nothing scarier for developers than fixing problems that they have no idea about where to start — because their microservice is just a piece of a larger puzzle composed by different other microservices, databases, and distributed systems. Though observability is the right answer for scenarios like this the reality is that it might not be in place yet to effectively make the difference.
The internet is not running in short supply of resources that explain in detail what this new buzzword called observability is about and why you need it so hard. However, very few explain how to implement it as quickly as possible to be used in troubleshooting scenarios like the aforementioned. This talk will show in 40 minutes how to implement observability from scratch using Elastic APM and how to leverage logs, metrics, and traces to identify problems in microservices.