A presentation at Abstractions II in in Pittsburgh, PA, USA by Matt Williams
Tracing the Black Box of Serverless matt williams – evangelist @ datadog
5 minutes after we first met
How I use this
It works great
No errors… Data shows up where its supposed to
But it feels slow
How do I find out where the problem is…
…when there is no problem
Three pillars of observability • Metrics • Spotting trends/patterns • Send alerts • Catch known unknowns • Logs • Traces
Relevant metrics to help me • Total ounces eaten • Performance timings for each function
Three pillars of observability • Metrics • Logs • Structured & easy to grep • Find the details of an event • Catch the unknown unknowns • Traces
Relevant log lines for me • Function entry and exit timestamps, total time • What is being collected by each step
Three pillars of observability • Metrics • Logs • Traces • Request-based • Follow activity across function and service • Answers “Where?” and “How long?”
But I want more visibility into what is going on
Time wasted getting static values
But I need more data
Matt Williams Evangelist at Datadog @technovangelist mattw@datadoghq.com youtube.com/technovangelist
Serverless is often a black box. You write functions then they are invoked on a platform on which you have no real visibility. Sure, there are metrics and logs, but finding the bottlenecks is a huge challenge. Understanding how a function relates to other resources is just as tough. On AWS, a tool that solves these problems is X-Ray. In this talk, I will show you how you can use X-Ray to:
troubleshoot the more ornery issues with Lambda
get a better understanding of how functions are invoked
work with all 3 pillars of observability to get a complete view of your serverless infrastructure
Here’s what was said about this presentation on social media.
for free. You
can too.