BuzzFeed Tech Talk Primetime: Monitoring and Observability
from Thursday, April 5, 2018 5:30PM to Thursday, April 5, 2018 8:00PM
Ticket Information
Event Details
BuzzFeed Tech Talks is BuzzFeed's speaker series where we gather with the rest of the NYC tech community and discuss topics that matter to all of us. The theme of this event is about monitoring and observability. When you system goes down, what do you do? Do you panic and toss your computer at the wall? Do you start guessing what's wrong? Or, do you use your system's metrics and dashboards to help yo figure out what the issue is? What, you don't have metrics and dashboards set up? At BuzzFeed, we live for having charts and instrumentation that can tell us immediately what the cause of a production is at a glance or as a result of a very concise on-call page. Come hear about how the NYC community sets up their system monitoring to avoid waking up in the middle of the night! PRESENTATIONS: Dashboards in Action at Datadog At Datadog we're in the dashboard business. We use dashboards to watch everything from the mythical to the mundane. A curated set of dashboards is the first thing we see when we walk into the office and the last thing we notice when we leave. This talk will give an inside look at the uses of dashboards at Datadog, including our custom-built solution for avoiding long lines at the restrooms, and our process for monitoring our own infrastructure needs. You'll walk away with tips from the source about how to corral your data into clear, actionable dashboards that foreground the most important metrics for your organization. Magic numbers in monitoring Setting up low-noise, actionable metrics is challenging, but there are times that the right graph can provide immediate answers to an incident or issue. In production, even the best-laid plans can go wrong as failures or unexpected issues occur. Kenton will discuss a few tricky situations where monitoring made a world of difference and a few reoccurring themes. Day to day trends When an incident occurs in production, we rely on metrics and logs to assist us in troubleshooting. While these metrics are crucial in the moment, viewing them on a daily basis can help us understand application behavior. This talk is about identifying negative trends early and becoming proactive in resolving the issue. Don't believe your lying eyes, and tags As a viral lab gone big BuzzFeed is concerned about downtime and we watch things like a hawk using DataDog. This talk is how one incident led us down the rabbit hole of tags and how to configure them to avoid artifacts in our data even as we `change the oil while going 100 mph`. SPEAKERS: Mark Hintz is a Software Engineer at Datadog, working on the dataviz team. His team focuses on making Datadog visualizations as fast, intuitive, and beautiful as they can be. Before Datadog, he worked at the visualization design studios Periscopic and Interactive Things. He holds a Master’s in interaction design from the Zurich University of the Arts and a Bachelor’s in economics from Reed College. Kenton Jacobsen is a Director of Engineering for Vogue, Glamour, and GQ at Condé Nast where he leads a team of software developers. Before this, Kenton led engineering for the viral news website, TheBlaze.com. In his free time, Kenton reads philosophy, rock climbs, and keeps up on Internet memes. Felicia Cippoletti is a Site Reliability Engineer at BuzzFeed, supporting MySQL and Redis. Previous to BuzzFeed, she worked with MySQL databases at Macys.com. Dan Meruelo is a Senior Site Reliability Engineer at BuzzFeed, working on the infrastructure team. His team focuses on creating secure, robust, and resilient systems at BuzzFeed. He built his first bicycle from scratch at age eight with a nutcracker and butter knife, and has been building things, big and small, ever since. He holds a Bachelor's in engineering from the Fu Foundation School of Engineering and Applied Science at Columbia University.
BuzzFeed HQ
111 East 18th Street New York NY USType: Eventbrite