Usenix LISA 2013: DevOps in every timeslot

There is a devops-related talk in every hour of this year's Usenix LISA conference. Usenix LISA Is a general conference with many tracks going on at any time. A little analysis finds there is always at least one DevOps related talk (usually more than one). This is very impressive. The problem, however, is that many of the talk titles don't make this clear. No worries, I've done the research for you.

[I apologize in advance for any typo or errors. Please report any problems in the comments. The conference website has the latest information. Other lists of presentations: Programming, Unix/Linux administration technical skills, Cloud Computing, and Women at Usenix LISA.]


Usenix LISA 2013 Presentations about DevOps:

Sunday, Nov 3, 2013:

Learn how Google SRE build and run huge systems. Google SREs were doing DevOps for years before the term "DevOps" was coined. John Looney is a Google SRE that created training programs inside Google to train more SREs. This is a "don't miss" session.

This workshop is an annual event at Usenix LISA. In past years this is where CfEngine was debated and Puppet was born. The last one had excellent presentations and I learned a lot. I see forsee good things here this year too.

A half-day workshop on running "Game Day" exercises. That like a firedrill for your service: take down random machines and make sure everything keeps running.

Monday, Nov 4, 2013:

  • 9am-5pm:
  • Cloud
  • Format: Workshop

While the title is positively vague, the description is clear that this is for people that run private clouds (or want to).

Tuesday, Nov 5, 2013:

Realizing that DevOps is all about working as a team, I've rewritten this half-day training to be all examples from working in DevOps organizations. People from traditional enterprise orgs will still learn a lot.

  • 1:30am-5pm:
  • Evil Genius 101
  • Thomas A. Limoncelli, Stack Exchange
  • Format: Half Day Training Class

First time at LISA! The first half of this class is all about applying DevOps techniques in the enterprise and everywhere. The second half is sneaky underhanded ways to convince people to adopt your ideas.

"You want to innovate: deploy new technologies such as configuration management (CFEngine, Puppet, Chef), a wiki, or standardized configurations. Your coworkers don't want change. They like it the way things are. Therefore, they consider you evil. However you aren't evil; you just want to make things better."

Nothing says "improved devops flow" like Continuous Integration. Jenkins is an amazing open source product that will let you apply CI to building software as well many other sysadmin tools (I use something like Jenkins to push DNS updates!)

If you remember Ben Rockwood's 2010 keynote he said ITIL has tons of great ideas that every DevOps engineer should know.

  • 9am-5pm:
  • Metrics
  • Format: Workshop

Brendan Gregg is a metrics god! I want to attend this just to bask in his monitorific glow!

Get your agile/devops flow on!

Wednesday, Nov 6, 2013:

What better way to learn Chef than directly from Opscode!

DevOps diva Mandi Walls talking about what skills you'll need in the future. If you want to learn about DevOps culture there are few better to learn from.

Adele will share what she's learned from running "game day" disaster recovery firedrills in IT shops. If you're familiar with the DevOps "anti-fragile" concept, this will be good training for you.

Caskey knows more about monitoring than ... everyone. His systems monitor more elements than he can say in public. Believe me. See this. You will want to apply everything he says to your monitoring system no matter how big or small.

  • 2:00 p.m.-3:30 p.m.
  • User Space
  • Noah Zoschke, Sr. Platform Engineer, Heroku
  • Format: Invited Talk

While the title is positively vague, the description is clear that this is about Heroku's innovative (dare I say "radical"?) ideas about the layering problems in cloud and hosted computing and how to solve them.

I'm biased because I shepherded this paper. Check it out: Google ran RedHat on its zillions-they-won't-say-how-many computers. Marc and his team was charged with converting them all to Debian. You'd think they'd reload them one at a time. No, for reasons he'll explain that wasn't permitted. So instead they changed one file at a time until the machines were running Debian. Yes. Really. It's an amazing story. There are lessons to be learned here whether or not you ever think you'll be in this situation. You will learn more about testing and, heck, "being careful", in these 30 minutes than in the rest of your career.

There are 3 other talks at the same time. All are great DevOps talks (storage performance, key-value at large scale, hadoop) but I'm not going to describe them because you should all see Marc's talk. (And OMG those other three talks are awesome DevOps-y goodness. Buy a clone. Send them to see the other talks. I mean it.)

Remember earlier when I said Caskey knows more about monitoring that anyone I know. I forgot about Theo. Theo is tied with Caskey.

Thursday, Nov 7, 2013:

Hilary is the data scienist that founded bit.ly. You definitely want to hear her speak. She'll keep it short. (ha ha, get it? Gosh, I bet she's never heard that before.)

  • 11:00 p.m.-12:30 p.m.
  • Women in Advanced Computing
  • Moderator: Rikki Endsley, USENIX Association, Panelists: Amy Rich, Mozilla Corporation; Deanna McNeil, Learning Tree International; Amy Forinash
  • Format: Panel

Last year the men that attended this panel learned more than the women. Don't miss it!

DevOps is LEAN applied to system administration. Ben's keynote kicked ass at LISA 2010. I look forward to seeing this presentation.

If you missed Brendan Gregg's workshop, you won't want to miss his talk about "flame graphs".

SDN is way cool.

Friday, Nov 9, 2013:

This is my "Ask Me Anything" session on Time Management. Last year the turn-out was huge. I'm not saying that to brag; I'm just saying you should come early because last year they turned away people.

What better way to learn Puppet Enterprise than directly from Puppet Labs!

Clay and Ed manage Macs like DevOps manage massive web services. The difference is that Clay and Ed probably manage more computers than most cloud services dream of having. If you manage Macs, don't miss this!

One of the most difficult things about running large services is I/O. I look forward to this talk.

DevOps applied to security. Awesome.

This talk will include a lot about Omega. If you have heard of Mesos, this is what they're trying to clone.

Todd is one of the funniest and best managers at Google. This talk will include his take on DevOps and how to end "toil".

Posted by Tom Limoncelli in ConferencesDevOpsLISA

2 TrackBacks

TrackBack URL: http://everythingsysadmin.com/cgi-bin/mt-tb.cgi/1645

DevOps at LISA from Everything Sysadmin on September 23, 2013 11:20 AM

A shout out to the conference planning committee of Usenix LISA this year. Narayan and Skaar did a great job! The amount of DevOps content is unbelievable. All 6 days have DevOps content that I want to attend from 9am... Read More

Someone asked me the other day if I had a "secret of my success". They didn't believe that I got this far on my good looks. (ha ha ha). For most of my career I've been on teams of people... Read More

Leave a comment

 
  • LISA16