The ZFS filesystem provides many powerful features such as built-in software RAID, the ability to self-heal from data corruption, copy-on-write, low-overhead snapshots, support for multiple boot environments, and more. This presentation provides a basic overview of these features and introduces the ZFS management tools which are incorporated into the graphical user interfaces of FreeNAS (an open source storage system) and PC-BSD (an open source desktop/server).
Dru Lavigne is the lead documentation writer for the PC-BSD and FreeNAS projects as well as a doc committer for the FreeBSD Project. She is author of BSD Hacks, The Best of FreeBSD Basics, and The Definitive Guide to PC-BSD. She is founder and current Chair of the BSD Certification Group Inc., a non-profit organization with a mission to create the standard for certifying BSD system administrators, and serves on the Board of the FreeBSD Foundation.
Around the world, people are enjoying billions of hours per month of streaming movies and TV shows from Netflix. In fact, nearly one-third of the Internet traffic in North America originates from Netflix servers, more than any other single source. The engine of this traffic is the OpenConnect network, a single-purpose content delivery network powered by FreeBSD 10. This talk will explore the architecture of the OpenConnect network, the key role that FreeBSD plays in its success, and a peek at what the future holds for FreeBSD at Netflix.
Scott Long lives near Boulder, CO, and has been working for Netflix since May 2012. He has been a user of BSD since 1992 and a FreeBSD committer since 2000. Scott was previously employed by Yahoo and Adaptec, and has worked extensively in device drivers and the CAM storage subsystem of FreeBSD. He holds a BS degree in Aviation, and enjoys flying airplanes and hiking in the mountains with his family.
Michael W. Lucas
What makes the BSD operating systems special? Why should you deploy your applications on BSD? Why does the BSD community keep growing, and why do Linux sites like DistroWatch say that BSD is where the interesting development work is happening? We'll cover the not-so-obvious reasons why BSD still stands tall after almost 40 years.
Michael W Lucas is a systems and network administrator with over 20 years of BSD experience. He's the author of the classic BSD systems administration books "Absolute FreeBSD" and "Absolute OpenBSD," as well as several other books on encryption, networking, security, and systems administration.
As the network architect for an independent Michigan telecommunications firm, Lucas runs life-sustaining systems through BSD servers. And sleeps well at night. Or during the day. Whenever.
Many network devices currently use BSD as their base and expose that in useful and interesting ways: Juniper, Netscaler, and Force 10 to name a few.
By understanding the underlying OS you can make better use of these machines to provide visibility into your network, troubleshoot better and faster, and increase uptime. You can use BSD tools to automate tasks, to perform packet captures, and to better understand your network.
BSD is well suited to the networker's way of thinking: logical, sane defaults, most of it works as expected, and few knobs. This allows you to build and maintain useful tools. Utility servers, firewalls, network access control systems, netflow systems, taps, and more depending on your knowledge and imagination. And all this knowledge is portable back to commercial systems. Secure, functional, and free, it is your networking LEGO set.
This presentation will present starting points for all of the above and an overview of how I have used BSD throughout my professional life to solve problems faster and better. While advising on how to get started and possible projects for improving and automating your network, this talk will also present ideas on how the BSDs can be improved to help networkers take better advantage of the amazing tools there.
Ray learned Unix and networking starting in an Air Force weather station more years ago than he cares to think about by spending every possible minute messing around with the systems. He currently works at Verizon Terremark as part of their Netops team running networks in world class data centers.
He has spent more than 15 years using his BSD knowledge to make networks of various sizes and types run faster, better, and more securely. His second favorite poem is about spanning trees.
Any large codebase such as a modern, general-purpose operating system has a process by which bug fixes, incremental improvements, and dramatic new ideas become part of the released system. NetBSD is an operating system which since 1993 has focused on clean code, a well-integrated system, and providing a stable research platform upon which great things can be built, and it has a particular approach to the problem of providing stable and tested releases. This approach is complicated by the number of targets that are built: 55 for NetBSD 5.x, 60 for NetBSD 6.x, 66 (and counting!) for NetBSD-current, which will become NetBSD 7 later this year.
In this talk, we will discuss the NetBSD approach to releases. We will cover how NetBSD can be built on almost any POSIX system, and how we have used this to create an autobuilder that builds releases of NetBSD for all supported platforms multiple times per day. We will discuss what's working for NetBSD, what's not working, and where we would like to go in the future. We will also touch on details like: supported lifetime of releases, testing methodologies, and the finer points of maintaining the quality of a world-class operating system with no paid staff.
Jeff has been a member of the NetBSD Release Engineering team since 2005, and a user since he was looking for a new OS for his Amiga in 1994. He's also part of the NetBSD admins team that keeps the build cluster and other systems running. He's run NetBSD on at least 7 different processor architectures, and lately has been enjoying running on low-end hobbyist ARM boards such as Raspberry Pi and BeagleBone.
A quick introduction to the installation and configuration of Postgres on FreeBSD and ZFS with a demo of using CitiBike data collected every minute since May 28, 2013.
The talk will cover:
Andrew Wong, Software Engineer for AppNexus, previously worked at Viggle and Gilt Groupe in NYC. The last three years I've worked on designing and implementing Data Warehouse loops with an emphasis on data freshness. Current working on the data delivery system for Real-Time Bidding (RTB) and optimizing/rationalizing databases. Still in search of why Ike thinks man sections 1-7 are a waste of space.