Event Details

At Datadog, our teams work together across the company–and the globe–to solve complex issues for developers, IT operations teams and business users in the cloud age.

In our event series, Datadog Tech Talks, we hear directly from industry leaders on the innovative, pragmatic projects they’re working on, as well as examples of real problems they’re solving for our customers.

This time we will be featuring two talks. For more information, please scroll down and check out the Talk Descriptions below.

"Random Load Balancing is Uneven"
"Using Open Source Tools for Vulnerability Detection"

REMINDER:

* Please bring your ID for entry to the building

* Please mind the legal age requirement - you must be 18 y.o

Agenda

Friday

2pm

Welcome + Networking Reception

Gather at Marché at 2pm for a welcome and networking reception. Meet colleagues and exchange ideas.

3pm

Opening Keynote: “Never Ending Game”

At 3pm we'll kick off proceedings with a keynote from art director Callie Moore entitled “Never Ending Game.”

4pm

Roundtable: “Protein Interaction Networks”

At 4pm, join a roundtable discussion on protein interaction networks with panelists Nell Richardson, Hugo Hitchcock, and moderator Ernest Arnaud.

Saturday

10am

Continental Breakfast

We return at 10am Saturday morning with a second networking session over breakfast.

11am

Ruby Coast: “Feeling Good About Feeling Good”

Join Ruby Coast for an interactive workshop session open to all.

12pm

Ann Tjader: “Here on Earth”

At noon, author and speaker and Ann Tjader will present important lessons gained from her third novel, “Here on Earth.”

Meet the Speakers

Evan Jones

Staff engineer

Evan has been a software engineer at Datadog for 4 years, working on a few different data storage projects. He currently works on Key/Value storage.

Janani Rajadurai

Software engineer

Janani is currently a Software Engineer on the Continuous Security team at Datadog. She is passionate about the intersection of Software Engineering, Security and AI as well as the utilization of tech for good.

Talk Descriptions

"Random Load Balancing is Uneven"

Speaker: Evan Jones

If you want to distribute data across multiple machines, we often use hashing. We usually assume this evenly distributes the items. Unfortunately, that is not true. In this talk, I will describe how this occurred in a real system, briefly discuss the probability theory behind it, then describe how this applies to other software systems.

"Using Open Source Tools for Vulnerability Detection"

Speaker: Janani Rajadurai

The ability to detect CVEs on our systems is reliant on accurately enumerating the libraries, packages, and dependencies that exist on those systems. Vulnerability scanners often use package manifest enumeration, causing them to miss embedded packages that can be detected by binary enumeration.

Through this talk, we will discuss how we leverage open source tools that are built using binary enumeration for package detection as well as specific enhancements that have been added to these tools. These tools allow us to create enriched SBOMs (Software Bill of Materials), that more accurately represent our software composition, which increases our ability to detect and report CVEs that exist on our systems. This allows us as a business to quickly patch our vulnerabilities as well as increase transparency towards our customers.

Agenda

6:00 PM - 6:30 PM

Reception and Arrival

6:30 PM - 7:00 PM

"Random Load Balancing is Uneven" with Evan Jones

7:10 PM - 7:40 PM

"Using Open Source Tools for Vulnerability Detection" with Janani Rajadurai

7:45 PM - 8:30 PM

Networking and Closing

Talk Descriptions

First Talk - Taming Alerting Fatigue: How Datadog Reacts to Unstable Data Sources

Speakers: Brice Figureau & Alex Olivier
At Datadog, we build and run large data pipelines, ingesting tens of GBs per second, to store metrics and events, and allow customers to be alerted based on these signals. But when data sources are disturbed, such a system is inevitably at risk of sending false positive notifications. Minimizing these is critical for an observability platform to be trusted by engineers, who could start ignoring real alerts that could prevent a customer facing incident if these are too frequent.

In this talk, we will explore the internal monitoring system we have created to detect such issues and minimize alert fatigue. We will cover the design and architecture of the system as well as the different trade-offs we had to make to ensure that important alerts wouldn’t be skipped. At the end we will cover some of the improvements and new features we have in the roadmap to continue making Datadog a trusted source for our customers.

Second Talk - Distributing Config Data at Scale in Datadog’s Event Platform

Speakers: Michelle Evans Bouchet & Jérémie Girault
The Event Platform is a central component of Datadog, which powers the backend of several Datadog products such as Logs, Traces and Network Performance Monitoring, ingesting tens of millions of events per second which represents hundreds of GBs per second.

In order to be able to ingest that data reliably, we are running thousands of containers in Kubernetes which need to be able to access fresh configuration data 24/7. Whenever a user changes a setting in the UI, this needs to be propagated to all relevant containers in real-time. For this, we have conceived a system to distribute this configuration to our workers in a reliable and eventually consistent manner.

In this talk we'll describe our initial scaling challenges, the design of the current system to address those, its architecture, and what we’ve planned next to improve it. We’ll also cover how this component became our main way to serve configuration to workloads in the Event Platform with over 10 teams using it.

Agenda

Learn More About #DatadogLife

• We provide many opportunities to innovate and solve complex, real-world problems and are rapidly growing

• We are on the search for the best and brightest recruits from around the world and want to get to know you
• We also want know know what sort of professional and intellectual environment you need to feel fulfilled

Want to know what #DatadogLife is really like? Hear from Datadogs around the world as they share their impact on our growth and what makes Datadog a great place to work!

Opportunities

Product Manager - Observability Products

Given the resilience of cloud technologies and importance placed today in digital operations and agility, Datadog continues to innovate and is well positioned for the long term. Partnering closely with our customers and internal teams of engineers, designers, marketers, and sales, you will be part of defining the future of a rapidly growing business.

⟶ Check out this opportunity

Product Manager - Application Security

The Application Security is the newest product of the Cloud Security Platform. As a Product Manager, you will shape a product helping change how cloud-native organizations monitor and protect their web-applications and APIs. Together with your team, you will build solutions that intelligently detect threats, correlate observability information, and ultimately help customers remediate and stop attacks.

⟶ Check out this opportunity

Product Manager - Real User Monitoring

Together with your engineering team, you will define the future of this fast-growing business. Specifically, as a Product Manager in the RUM team, your work will help make RUM a best-in-class troubleshooting product for more platforms and technologies. Your role will be to ensure that the product is configurable, yet easy to adopt at scale, and deeply integrated with other Datadog products. Your decisions will have a deep impact on both the product and the day-to-day job of our customers.

⟶ Check out this opportunity

RSVPs Closed

Modern Monitoring & Security | Join the Pack!

Read about our Privacy Policy and our Terms of Use.