Math is hard and intimidating, but it doesn't have to be! This talk covers advanced topics in Cryptography, condensed into the most essential information. The goal being to equip you, a kickass coder or admin, with some of the theory behind the protocols of past, present, and future.
Imagine having to upgrade 100,000 servers to install a kernel patch. In this talk, you’ll hear how the Netflix Operating System team uses automation to keep a common Linux base image up to date, and how our application teams deploy their applications on this image, always picking up the latest. We’ll talk about how we insulate developers from the operating system and tailor the upstream Linux distribution for large scale cloud deployment. We’ll also discuss best practices such as using immutable infrastructure and designing for failure to insure service availability.
"We will encourage you to develop the three great virtues of a programmer: laziness, impatience, and hubris." – Larry Wall.
DevOps engineers, if anything, are the epitome of the three great virtues of programmers. Since the advent of DevOps in 2009, engineers created a plethora of tools and systems to automate every aspect of the software development and production lifecycle.
This presentation will dive into the answer of: what happens when a DevOps engineer uses their tradecraft in their home-life?
DNSSEC DANE and DNSCrypt - The Internet hasn't seen this kind of improvement since Jon Postel's original DNS RFCs. DNSSEC brings authentication of information served by authoritative servers thus the dawn of the believable Internet. From this, the emergence of DANE or DNS-based Authentication of Named Entities. This adds confidentiality to believability. Wrap all that in DNSCrypt and you have a confidential, verifiable and decentralized naming service. While not all Internet infrastructure has left the era of blind trust, the names we depend on are now things we can verify.
The movement for free exchange of software stemmed initially as a backlash against restrictive licensing regimes. As a response, communities formed their own licensing rules that used that restrictive system of control as a mechanism to liberate us. Thirty years later, this licensing system of open sharing, which is the bedrock of FLOSS, shows cracks, particularly with the advent of for-profit companies participating heavily. This talk explores the myriad of issues in the licensing structure for FLOSS, and considers what community-oriented actions are needed to address these problems.
When we in technology communicate with people who are less technical than we are, we can sound a lot like aliens. We're speaking a foreign language, but thankfully we also know the language of the natives. All we need to do is remember when and how to use it.
This talk explains the transition process the Logz.io R&D team underwent to move to Continuous Deployment using Kubernetes - existing challenges, the reasons Kubernetes was chosen for Docker orchestration, the implementation process, and the cultural change needed for making the shift.
eBPF has gained popularity in low overhead tracing and data aggregation used to understand inner workings of the Linux kernel. The bcc tools based on eBPF work only on x86 machines where they are running locally (host and target are the same). This requires llvm libraries, Python and the kernel sources all installed on the local machine, before eBPF can even work. This installation is easy and readily available as packages on x86 however it is painful to do for arm64. Joel addresses these challenges with a novel solution with a new tool (bpfd) he created to remotely use eBPF.
The ELK was the first system I implemented at Nativo in Mar '14. Elastic Stack is now a critical system. We will not deploy to production without the visibility that the Elastic Stack provides. Elastic Stack is ingesting over 30 Application generating 150M documents and consuming 225GB daily across four AWS regions
Prometheus is an open source metrics collection and alerting system. It can monitor anything that exposes a /metrics endpoint! A single small instance of Prometheus can monitor tens of thousands of metrics from hundreds of applications and break no sweat.
In this walk, we will demonstrate the power of white-box monitoring and tell why you should instrument your microservices and even your daemons with metrics endpoints. We will hands-on instrument a real-world application that's been running in production and write alerting for some of its metrics, live on the stage.