SF Reliability Engineering - September Talks

Doors at 6pm and ID required in the lobby. We're starting the talks at 6:30pm. Food and drinks provided.

--

We're always looking for 20-30 minute technical talks (and 5-8 minute lightning talks) relating to the very broad field of Reliability Engineering. Get in touch with the organizers if you'd like to present!

Important check-in details for Twitter:

1. Bring your photo ID. This is required by building security during the check-in process.

2. We will begin checking people in at 6:00pm. Check-in closes at 6:30pm, so please arrive on time!

(If you arrive after 6:30pm, you may message the organizers on Meetup.com and we will do our best to get you into the event. However, please note that we are unable to check-in any attendees after 6:45pm)

--

Twitter Developer Community Code of Conduct:

Events hosted at Twitter are covered under the Code of Conduct: https://developer.twitter.com/en/community/code-of-conduct.html

--

Talk #1: Making Service Launches Boring with Distributed Tracing

Description: Launching a new service doesn’t need to be stressful. Thinking about (and implementing) Distributed Tracing and Observability from the onset can make deployments seem routine and boring. In this talk, we will talk about how LightStep launched several new services without downtime or data loss. We’ll cover conventions and standards we follow internally and how Distributed Tracing fits into the process. Deploying a new service to prod is only the first step; optimizations and bug fixes are what guarantee our job security. We’ll discuss ways Distributed Tracing can help us quickly find and fix bottlenecks.

Bio: Karthik is a Software Engineer at LightStep, building tools to make on-call life a little better. His interests are in software performance, reliability and observability. Previously, he has worked on several monolith-to-microservice migrations and his most recent project involved building aggregate trace analysis at scale.

Talk #2: strace: A case study in troubleshooting

Description: When problems occur only in production; when your IDE can’t locate the bug; when closed source vendor code keeps crashing; and when complex inter-process interactions confound the best troubleshooting efforts... you can take Linux’s powerful strace command out of the toolbox. Just as tcpdump sees everything on the network, strace sees everything that a process does with the outside world (kernel, network, disk, etc.). See how the power of strace can be used to peel back all the layers of complexity and discover what’s actually happening at the lowest level.

Bio: Max Euston leads a team of SREs at WePay to ensure the smooth operation of online card payments behind some of the most recognized brands on the web like GoFundMe and Meetup. He is a classical hacker who grew up with technology. Prior to the commercial realization of cloud computing, Max build a fully-automated, data-driven compute environment for Prudential’s entire web presence. He is deeply passionate about both the technical and human aspects of IT.

--