Tag Archives: talks

OpenStack Summit Preview: Elastic Recheck

With OpenStack summit only a few days away, I’ve been preparing materials for my Elastic Recheck talk. Elastic Recheck is a system that we built over the last 7 months to help us data mine failures in test results in the OpenStack test system to find patterns.

The Problem

OpenStack is a complicated system, with lots of components working in an asynchronous way. This means that small timing changes can often expose some interesting issues. This is especially true in an environment like the upstream gate where we are running tests in parallel.

A good example of this is a currently open bug. If you run a security group list against all tenants at the same time someone is deleting a security group, the listing returns a 404. This is because of a nesting behavior in the list, which includes running a db get over all the items in the list to get additional details. There is an exposure window there where a security group is in the list, it’s deleted by another user, then we go back to get it, and it fails. That failure currently propagates a set of exceptions which become a 404 to the end user. Which is totally unexpected.

That window seems really small, right? Like it never could actually happen. Well, in the gate, even with only 2 – 4 API calls happening at a time, we see this 7 times a day:

Security Group Listing

The Solution

Starting during the Havana RC phase, we started turning this into a search problem. Using logstash and elastic search on the back end, we find fingerprints for known bugs. These fingerprints are queries that will give us back only test runs which seem to have failed on this particular bug.

The system includes real time reporting to Gerrit and IRC when we detect that a job failed with a known issue, and bulk reporting every 30 minutes to let us understand trends and classification rates.

Elastic Recheck Flow

Overall this has been a huge boon towards really identifying some of the key issues we expose during normal testing. What’s also been really interesting is having a system like this impacts the way that people write core project code, so that errors are more uniquely discoverable. Which is a win not only for our detection, but for debugging OpenStack in a production environment.

Learning More

If you are going to be in Atlanta, and would like to know more, you’ll have lots of opportunities.

My summit talk, which is going to be overview intended for people that want to learn more about the project and technique.

Elastic Recheck – Tools for Finding Race Conditions in OpenStack
Date: Thursday, May 15th
Time: 2:20pm
Room: B206
Track: Related Open Source Projects

We’ll also be doing a design summit session for people that are interested in contributing to the project, and helping us set priorities for the next cycle. Wed, 9:50am in the Infrastructure Track.

Also, feel free to find me anywhere to chat about Elastic Recheck. I’m always happy to talk about it, especially if you are interested in getting involved in the effort.

I believe the summit talk will be recorded, and I’ll post links to the video once it’s online for people that can’t make it to Atlanta.

Drupal Talk Roundup

A couple weeks ago I gave the first MHVLUG talk of the year on the work I’ve done with Drupal for the Poughkeepsie Farm Project, Mid-Hudson Astronomical Association, and MHVLUG. I think it went well, but it’s sometimes hard to tell, especially because I’ve been trying out some new approaches on talks, so feedback comes in different ways.

When I was building the Android presentation for CPOSC, I realized I was building a generic Android tutorial, and I stopped myself. Why would anyone want to hear me give that presentation? You can get that all over the internets, on the youtubes and vimeos of the world. So, I thought long and hard about what I could uniquely bring to the table. The answer was pretty simple, talk about the things that I racked my brain on, in the context that I ran into them when building my application.

This drupal talk was much the same. Two days before the talk, while doing a dry run at home, I realized that I had a tutorial that was better covered by the internets. So, I started over, and edited heavily, and managed to produce a personal narrative of working with the PFP that showed the nuts and bolts of Drupal along the way. It was a much better presentation. It also made it easier to put a call for action in the presentation, for people to get involved with local non profits.

Because this was a narrative, and not a tutorial, the audience interaction is very different. People stop and ask clarification during a tutorial, but questions during a story are often considered rude. This led to a very distinct boundary between talk, running just about an hour, and questions, which carried for 30 minutes after that. Without questions during the talk, it’s a little harder to tell if people are engaged. I do have a vivid image in my head of looking out out on the audience and even the laptop folks were all eyes up and watching, so I think I succeeded there. The 30 minutes of questions, coming from at least 8 different people, helped reinforce that.

I also made a few mechanical changes to the presentation in Open Office, which others might find interesting. I’ve moved to using Fade Smoothly (Fast) for slide transitions, and Fade In (Fast) for element animation. Getting rid of the hard appear makes the whole thing feel more fluid, and as long as it’s fast, it doesn’t get in the way.

I also realized that when you put up a slide you feel like it should get your time, but sometimes there is nothing really to add. I had a brief walk through of drupal installation. During dry run I realized I spent way too much time talking about slides with relatively little talk worthly content. To keep myself honest during the talk I made those slides automatically advance after 5 or 10 seconds (depending on complexity). That meant I got all of that on screen, but capped it to 40 seconds, and didn’t find myself saying something like “what more can I say about this slide”.

Overall I think things went quite well, and am now looking forward to the ACM talk tomorrow night.

Android Development Talk at Poughkeepsie ACM

As confirmed yesterday I’m going to be presenting my Android development talk from CPOSC (with a few tweaks for the local audience) at the Poughkeepsie Chapter of the ACM this month.

Mon, Jan 17 7:30 pm – 9:00 pm (Ulrich Room, Dyson Hall, Marist College ): Poughkeepsie ACM Talk: Solar System in your Pocket – Developing Android Applications

It started with a simple discussion after a local astronomy meeting trying to figure out which moons of Saturn we were looking at. This seemed like the perfect first Android application, building an astronomy simulator that would let me answer that question wherever I was. Little did I know that trying to do this would take me on a Journey through most of the major subsystems and interfaces in the Android SDK.This talk will take you along on that journey of writing your first Android application.

It will touch most of the major concepts involved in mobile development for Android, and many of the interfaces you’ll need to write you first application. Most importantly it will give you a list of things *not* to do when developing for the mobile space.

Sean Dague has been an open source software engineer in the IBM Linux Technology Center for the last 10 years. His spare time is split between the outdoors, amateur astronomy, and random bits of open source hacking. Dinner before the event at the Palace Diner at 6pm.

The talk will cover some of the basics of android development, focusing specifically on how it’s a different programming model from what you might be used to.  I use my Where is Io application, and what I learned along the way, as a roadmap for the talk.

Upcoming Talk: Building a Community Site with Drupal

On Wednedsay, January 5th, I’ll be giving the MHVLUG lecture on Drupal.  It’s been two years since I started poking at Drupal in order to overall the Poughkeepsie Farm Project website, and this talk is largely going to be about that experience, and some of the basic lessons I’ve learned along the way.  While there will be plenty of technical bits, covering basics of getting deep enough into a drupal project to make it interesting, there is also another interesting story about getting involved with non profits.

The talk is coming along nicely, and I’m quite excited for giving it next week.

Speaking Tonight in New Paltz

Tonight I’ll be giving the Mid Hudson Astronomical Association talk on the basics of celestial mechanics, and what I figured out to make my Where is Io application.  The talk picked up the poetic “Tracking the Movement of the Heavens” title somewhere along it’s development.

This is going to be a tough talk, given how varied our audience tends to be, but I’m hoping that I’ve got enough hooks to get people in early, and plenty of meat for people that want to know more. The practice run through after work last night went quite well, the narrative turned out to be stronger than I realized, and I got some good feedback that definitely makes this a better presentation.  Thanks to Mike, Jeremy, and Sean for that.

Eben Moglen’s invited talk at IBM Research

About 3 weeks ago Eben gave an invited talk at IBM Research about Copyleft Capitalism, which I’m kicking myself for not driving down to see in person. Fortunately it has been posted to Google Videos:

(direct link for those whose RSS viewer doesn’t do flash)

The talk weighs in at 91 minutes, and is really worth watching the whole thing. Unfortunately the audience isn’t miked, so there are some areas of silence which are a little awkward, but that doesn’t detract much from the talk. Eben is an incredibly engaging speaker, and weaves a wonderful narative around free software, and the software commons.

While there are so many interesting points that Eben makes, I found the most fascinating one to be about how creativity can not be contained and controlled. When the industry attempts to do so you get bursts of irrational creativity, such as Perl. Perl is about getting things done, and it does so by hook or crook, based on whatever made sense to go into the language during last saturday’s hack fest. While it disgusts many of the purists in language design, it shows that sometimes “just get it done” does work. Beauty isn’t everything. (Steve Yegge’s take on perl is also hillarious and dead on.)

This is one nugget of dozens in the talk. I think I’ll need to go back and watch the thing at least once more to make sure I’ve absorbed it all.

Dague’s Rule 47: if Eben Moglen is speaking anywhere within driving distance, I must travel to the talk.