|Saturday May 3rd, 2014|
|8:00am 9:00am||Registration and Welcome Coffee|
Brunswick Ball Room A
Brunswick Ball Room B
Brunswick Ball Room C
Brunswick Ball Room D
Devops Talks Track
Infrastructure Talks Track
Professional/Career Talks Track
General Topic Talks Track
|9:00am 9:40am||Choosing the Right Config Management Tool for Your Environment
|Enlightening Technical Leadership
|Step AFK: Practical Advice for Career Management
|Starting & Growing a LOPSA Chapter
|9:45am 10:25am||Code Review for Sys Admins
Elizabeth Krumbach Joseph
|Using Ansible to Fill the Gaps Left Over from Puppet and mCollective
|IT Careers Panel
|Power & Cooling for the Small Office Server Room…or Closet
|10:30am 11:00am||Morning Break|
|11:00am 11:40am||DevOps in a Windows World
|Git Hooks for Sys Admins: With Puppet Examples
|Backing Into Configuration Management
R. Marshall Webber
|Switching to Linux for…Switches?
|11:45am 12:25am||Operations Level Up
|The Stack at Stack Exchange
|Professional Topics Panel Discussion
Dave Parter and Philip Kizer
|Where MDM on iOS Leaves Off: It’s Not the Be All, End All, So You Have to Plan
|1:30pm 2:10pm||You are Not a DevOp
|Building an Elastic Interface on EC2 with Puppet
|Don’t Be That Guy
|Meraki Networking: How I Learned to Stop Worrying and Love the Cloud
|2:15pm 2:55pm||The Future of System Administration (and what you can do to prepare)
|Using Vagrant to Remove the Cost from Experiments
|Fundamental Techniques for Release Engineering in the Real World
|Presentation Skills for the Highly Technical Employee
R Marshall Webber
|3:00pm 3:30pm||Afternoon Break|
|3:30pm 4:10pm||Testing Your Tickets: How to Reduce Rework & Impress Your Customers at the Same Time
|Containers vs. Virtualization: An Introduction for the Impatient
R. Marshall Webber
|Tom’s Top 5 Time Management Tips
|4:15pm 4:50pm||Why Write a Security Plan?
|Packing It In: Images, Containers, and Config Management
|Book Preview: The Practice of Cloud Administration
|5:00pm 6:30pm||Garden State Ballroom
Keynote by Elizabeth Krumbach Joseph: Universal Design for Tech: Improving Gender Diversity in our Industry
This talk will briefly cover the concept of configuration management and how it can fit into your infrastructure. We will then look at the features and capabilities of popular solutions such as Puppet, Chef, and SaltStack among others to determine their strengths and weaknesses in various environments. My goal for the talk is to provide the attendees with the knowledge to pick the correct solution for their infrastructure. This is going to be a crucial part of the information technology conference.
Who are you? Why do you do what you do? By asking yourself these questions and paying careful attention to the answers that emerge, you can become a better leader. When you bring greater self-awareness to your technology work, your vision is expanded, your creativity is boosted, and you can collaborate more effectively with a wider range of people. This session is about ways to ask and answer fundamental questions that can increase self-awareness, with insights from psychology and neuroscience. We’ll explore the following areas in the context of technology work:
- Refining, switching and letting go of your mental models, which increases your mental fluidity
- Understanding your sense of abundance and scarcity, which helps you invest your time wisely
- Observing how you set and form expectations when you communicate, which helps you collaborate more effectively
- Identifying the polarities that contain contrasting perspectives, so that you can make better sense of opposing pieces of advice
Technical skills alone are not enough to help you advance your career or find that next job. This session will provide some very practical advice about the things you can do to help broaden your horizons and prepare you for the next step in your career. One warning: your boss doesn’t want you to attend this talk, your takeaway might just be that you should quit your job!
Want to start a LOPSA chapter? Want to grow your existing chapter? This is the talk for you. Mike will talk about his experiences with starting and growing LOPSA-ETENN, including handling sponsorships, finding speakers, and recruiting attendees.
I’m a member of the OpenStack Infrastructure team which is a geographically distributed team of systems administrators from several different companies who work together in public to maintain the infrastructure described at http://ci.openstack.org.
To achieve this, we use a code review system that leverages Gerrit as the interface for peer review and Jenkins to run some basic configuration and code syntax checking against our submissions. This allows us to maintain for code and config file integrity and gives us a nice platform so that our fellow systems administrators can comment on and improve solutions we come up with. We also use IRC, Etherpad and more for collaboration, which I will discuss.
Sometimes ssh in a for loop is a solution. There are tasks that are not repetitive and don’t need the infrastructure of a configuration management like puppet or chef. For those day to day system administration tasks, ansible is a great tool for streamlining your tasks. In this talk we’ll cover how to setup your machines to use ansible, introduce the various plugins for ansible and show how to do adhoc tasks with ansible. A brief introduction to playbooks at the end, to show how to extend ansible to be more like a full blown configuration management system.
John Boris is the Director for Information Systems for the Office of Catholic Education in the Archdiocese of Philadelphia. He currently manages the Student Information System for the High School System of the Diocese. He has been involved in System administration since 1986 with the Defense Logistics Agency setting up the agency’s first Electronic Contracting Web site and Bulletin Board System. His is a current member of the LOPSA Board of Directors and an original member of the Mentorship Core team.
This talk will explain the basics of powering servers, sizing backup power and cooling your equipment. Despite the fact that server consolidation and cloud based services have removed some of the local infrastructure from small businesses, there is still a need to ensure a reliable, fault tolerant environment that can maintain connectivity to both local and hosted resources.
DevOps on Windows isn’t as straight forward due to the fact that there isn’t as much tooling out there as there is for the Linux world. George Beech will talk about how they bring the philosophies and principles of DevOps to Windows at Stack Exchange. He’ll also talk about some of the tooling – both in-house and external – that they use. George will go over some tips, tricks and strategies to start bringing these DevOps ideals to Windows administration.
Using hooks to make your commits do all the work for you. Working in a group, it’s important that tasks are automated so that all the users can work simultaneously. Using hooks we’ll look at how to automate deployment. After covering the different hooks and how to write them, we’ll look at some examples with puppet. We’ll get the hooks to deploy puppet environments for us.
Unless you’re starting a company from scratch, building a new data center or standing up a complete cloud infrastructure, you’re going to have a rough time implementing Configuration Management in a legacy environment if you try to start with a blank slate. There are players, fiefdoms, and cruft that can wreck even the most determined effort to ‘Automate Everything.’
Don’t wreck yourself trying to move your organization to a CM and Lifecycle driven infrastructure! Come to this talk and learn:
- How to get everyone rowing in the same direction by starting with the inoffensive processes that everyone wants.
- Describing CM in terms that your internal customers want, and will come to demand.
- How to help your internal customers (and by extension, yourself) become a partner, rather than a competitor in determining their needs for Configuration Management.
Today’s network device operating systems are just like the way they have been for the last 20 years — a “black box” constructed under a closed development model, available only on a given vendor’s hardware. Ofttimes there is no API, or when there is, it is vendor-specific. This leads to poor automation ability, duplication of admin skill sets needed to configure and manage heterogeneous vendors’ equipment, and increased cost due to vendor lock-in.
Server folks have lived thru that era before – but the advent of commodity x86 hardware and the birth of open operating systems such as Linux dramatically changed the landscape. Can this same solution be applied to network equipment?
This talk will explore the advent of “white-box” switching (including routing and firewalling) that is now threatening to upset the dominant networking vendor applecart. It will especially focus on what one vendor, Cumulus Networks, is bringing to the table with their offering of a pure Linux distribution (Cumulus Linux) that is built to run at wire speed on “white-box” switches offered by a number of vendors, which have the same feature sets as switches offered by the likes of Cisco, Juniper, Arista, etc. (mainly because they all use the exact same merchant silicon chipsets.)
This session is designed to help junior and mid-level operations practitioners to develop their skills and become more effective in their jobs. Topics will focus on using open source tools to build more robust processes for building, deploying, and maintaining web infrastructures.
Topics will include:
- Tools for building operating system packages for your own deployments
- Working with software development tools, version control, code review
- Testing methods, configuration management tools
We know that the day-to-day jobs of Operations folks is very busy, and it can be very time consuming to sift through all the awesome open source projects available that can help make our jobs easier. Many junior operations folks know there is probably a better way to do a given task, and we’d like to help give them some suggestions.
At Stack Exchange, we run lean and mean serving up Stack Overflow, Server Fault, and other great Q&A sites. In this session, I’ll describe the architecture of the Stack Exchange network and how we serve six million people a day and serve up over 22 million page views daily. I’ll point out how we make our infrastructure as resilient and efficient as possible despite running on a limited amount of physical hardware. I’ll also describe why Windows Server, .NET, and SQL Server make sense for us.
This is an audience-participation discussion. The panel will have topics prepared if filler is needed, but the best format is for an attendee/participant to pose a question or raise an issue about their career, workplace, or a higher-level technical/architectural question about their site. The panel will ask follow-up questions to “diagnose” or “debug” the situation, to get to the root of the matter. Other attendee/participants will be asked for their input or ideas, and the panel will provide their advice/analysis.
Another way to understand this session is a more formalized/moderated “hallway track” with a focus on professional/career issues. An example of a technical issue that would be appropriate would be how to introduce change management to a site, or “is my site ready for change management”? An inappropriate topic would be asking specifics about a tool or technology.
Non-techs hear full device management when you say MDM but that is not the full picture on iOS. There is no scripting, automation, or VNC, and the control you have is dependent on a network connection, so you have to plan around the limitations. Compounding this is if the device is not set in Supervised mode then the user can remove the control you have at any time they want. Lets discuss the limitations and pitfalls to make the most out of an MDM solution.
With titles and groups like “DevOps Engineer”, “Director of DevOps”, “DevOps Team”, clearly, DevOps is a thing. But can you really be a DevOp? In this talk we’ll explore some of the history and meaning of DevOps, what it means to “do the DevOps” and how you an take the first steps toward transforming your organization into one that uses DevOps practices to delight your customers.
One of the more innovative uses of Amazon’s EC2 is the ability to support elastic workloads that can expand and contract resources in response to demand. However, developing an application with the intention to be “elastic” does not remove the need to manage and maintain the systems it will run on.
Additionally, machines need to be capable of being brought up and handling load in relatively no time at all.
This talk will explore building an elastic infrastructure on EC2 using Puppet. We will explore the differences of elastic infrastructure compared to traditional static infrastructure. In addition we will look at a method for automating the maintenance for keeping the machine image maintained for rapidly ready to deploy instances.
A talk on why attitudes matter. Perpetuating the stereotype of the BOFH may seem funny, but it is also career limiting. Sarcasm and snark have their places, but some people will think you have an attitude problem. Other attitudes such as the egotist and the information hoarder are just as negative. Using these defenses to keep others at bay so you can avoid working with others or be the keeper of the knowledge isn’t how to get ahead in business. The old ways of the sysadmin in a dark corner have given way to a new type of operations professional. These attitudes also perpetuate diversity problems, they interfere with our ability to be productive and useful to management, and they’re holding you back.
Meraki is a relatively new networking company, very recently acquired by Cisco. They produce a relatively unique style of mid-level networking equipment configured and managed entirely though a remotely hosted web interface. This style of management, and their implementation of it, introduces unique opportunities and possibilities, both positive and negative. This talk will cover my initial testing of their devices, our decision to start deploying them, and our experiences with deployment and management of networks.
System Administration is changing faster and more radically that at any time in its history. Like it or not, your job is probably going to be very different in 10 years (and maybe even in 5 years); come hear where the field is headed and what you should do if you want to stay on the leading edge of system administration. (Sneak preview of the recommendation: “stop being a sysadmin.”)
From time to time, the need arises to quickly explore, test, or validate a hypothesis concerning a configuration change, upgrade, or software installation . In some cases, this may mean the risk of completely destroying the system the test is being performed on. Over time, even in some cases where automation is adopted, the cost associated with doing such a test begins to dissuade teams from attempting such trial and error.
In this talk, I will be using Vagrant and some examples to show how quickly and cheaply norisknocost experimental environments can be built. I’ll share my experiences as a consultant for Puppet Labs helping clients overcome their avoidance of doing such testing, allowing them to embrace experimentation and promote creative problem solving. I’ll also demonstrate a few useful
plugins to Vagrant that help speed up the process further.
Although the fundamental goal of release engineering–to deliver high quality versions of software to end users—have more or less remained the same over the lifetime of the existence of the discipline, generally accepted best practices have evolved considerably, especially over the past several years. Certain techniques that were widely accepted and largely unquestioned as recently as 10 years ago have been largely abandoned in favor of various agile methodologies that encourage developers to “release early, release often.” I propose to give a talk that distills currently accepted release engineering techniques, drawn from various free and open-source software projects and other organizations that publish their release engineering methodologies, as well as from my own experiences as a Unix systems administrator who has often found himself thrust into a release-engineering
Topics I will go over (in the context of release engineering) include: software versioning, documentation, configuration management, release frequency, security, testing, monitoring, and continuous integration. As much as is realistically possible, I will try to remain software and OS-agnostic.
The aim is a to propose a set of “fundamental techniques for release engineering” that most users should follow, in the hope that others would benefit from the recommended decisions and, ultimately, advance the existing state of software engineering.
You’ve got wicked hacker skills, debug and compile kernel modules for breakfast, and leap stack-overflows in a single
bound. Yet when you try to share specific How-Tos, system designs or even policies with a non-technical audience or
even a room full of your peers, you still get a lot of confused looks and or outright boredom. Nobody is getting the
message. How do you bridge the divide between your knowledge and your audience’s need to acquire it?
Appropriate for every skill level of SysAdmin or DevOps, this densely packed presentation (what else?) will equip you
with easy game changers which anyone can use to get their message across in ways that are ‘sticky’ and stay with their
audience long after the presentation is over. Whether you are the most Senior Admin or the newbie, you’ll walk out of
this session armed with an arsenal of ideas for keeping your audience engaged while you pass your message to them.
For almost two years, Chris has been using Cucumber to test work his customers have requested via tickets. In this talk he will discuss this testing, his experiences with it, and why everyone should test their work, even if (and especially if) they’re short on time.
Linux Containers are the new hotness for the Cloud and SaaS, PaaS & IaaS. What about for conventional customers who
are just now accepting virtualization? How do Containers differ from Full and Para-Virtualization that VMware and Citrix
Xenserver have popularized and monetized?
Before you sign another expensive Virtualization management contract, come take a look at what Linux Containers are,
where they make sense from a maturity and security standpoint. A serious, unbiased look at whether adoption will save
your company VM management licensing costs, or just shift those costs to implementation of Containers.
Basic understanding of Full and Para-Virtualized systems is a recommended pre-requisite, though extensive experience
with large virtualization deployments is not.
Time Management Guru Tom Limoncelli will reveal his top 5 time management tips and take Q&A from the audience. Tom is the author of the book Time Management for System Administrators from O’Reilly and has been a sysadmin for more than 20 years.
Learn about the many different tools gaining momentum to manage system configurations. Hear about when you should think about configuration management tools and when it might be OK to just pack it into a re-usable image. We’ll cover the basics around Docker and Packer before diving into a full-stack example leveraging all three topics in harmony together.
The team of Limoncelli, Chalup and Hogan are at it again with their new book The Practice of Cloud Administration, due out in November 2014. In this session Tom Limoncelli will give a preview of the book, and talking about the book writing process.
Why Write a Security Plan?
If you’ve spent any amount of time in the systems administration field, you’ve probably known someone who has written a security plan, or perhaps you’ve even had the *privilege* of writing one yourself. We’ll briefly look at what goes into a security plan, but more importantly, examine some reasons why we spend hours filling out forms and templates and checklists. Checking the box is a good thing, right?”