Honeycomb Tech Talks 007: Serverless Observability
May 16, 2018
6:00PM - 8:30PM
Every month we host a loosely-themed set of 2-3 talks focused on topics related to operations, performance, debugging, and observability of modern web/saas/etc services.
Honeycomb Tech Talks is a recurring series hosted by Honeycomb. Everyone is welcome! <3
Event driven configuration with AWS Lambda
by Ryan Armstrong, Dropbox (SRE):
AWS Lambda and its serverless friends offer a robust means to automate dynamic configurations in both AWS native and third party services – like monitoring tools. We’ll explore one real-world example, how it exploded everywhere and how we eventually taught it good manners.
Serverless Gone Bad
by Eamonn O’Brien-Strain, Google:
Building on a serverless platform is great. You can concentrate on your business logic and leave all the lower levels of the stack for your serverless provider to worry about. But what if things go bad? What if your code has bugs? How do you debug it? How to you cope with the occasional errors, or performance degradations from the platform? How do you protect yourself against the new attack vectors that malicious actors can use against serverless systems? In this talk I’ll look at different types of serverless platforms (BaaS, FaaS, and event-driven) and talk about how developers can cope with various types of badness, and what providers of serverless systems are doing to make your life easier.
The Quest for Turn-key Serverless Agents
by Travis Redman, Honeycomb:
So you’ve got a serverless app generating logs and events – how do you get them in your tool of choice without a lot of steps? At Honeycomb, we adapted our server-based honeytail agent to run in AWS Lambda. In this talk we’ll explore what we learned along the way about Lambda, the AWS ecosystem, and packaging serverless agents for turn-key deployments.