• Home

  • Schedule

  • Speakers

  • Registration

  • Call For Proposals

  • Our Sponsors

  • Financial Aid

  • FAQ

  • Blog

  • Contact Us

  •  

    Sponsors
    Screen Shot 2016-07-15 at 8.37.57 PM
    Screen Shot 2016-07-15 at 8.37.57 PM
    Bloomberg
    Bloomberg
    HERE_Logo_RGB (1)
    HERE_Logo_RGB (1)
    new-twitter-logo-150x150_edited
    new-twitter-logo-150x150_edited
    Yelp
    Yelp
    sentry
    sentry
    Google
    Google
    Google
    Twist Bioscience
    Twist Bioscience
    Screen Shot 2016-08-05 at 3.09.22 PM_edited
    Screen Shot 2016-08-05 at 3.09.22 PM_edited
    cloudera_logo
    cloudera_logo
    sauce-labs200x200
    sauce-labs200x200
    hired-logo
    hired-logo
    PSF
    PSF
    shippo
    shippo
    twilio
    twilio
    Minted
    Minted
    Paypal-logo-20141
    Paypal-logo-20141
    anaconda-logo
    anaconda-logo
    microsoft-logo
    microsoft-logo
    eventbrite
    eventbrite
    Show More
    PyBay Connect
    • Meetup_square
    • White Twitter Icon

    Subscribe to PyBay Updates

    TOP

    Eben Freeman

    Bio

    Eben is a software engineer based in San Francisco. He's used Python to do math research and build email infrastructure, among other things. He likes pie, and rock climbing.

    Aug 20 4:30p - 5:20p, Fisher West
    Python tracing superpowers with systems tools

    Performant Python, Intermediate

    ​

    Description 

    Modern system tracers like SystemTap or Dtrace are incredibly powerful. If they're not part of your arsenal of techniques for analyzing Python code, you might be missing out. In this talk, we'll explore how these tools work, and how they can be used for dynamic, low-overhead analysis of unmodified Python programs.

    ​

    Abstract

    Maybe you want to profile your program, but it's running lots of C extension code and conventional profilers can't help you. Or maybe you're tracking down an emergent problem in a production system, but the logs are barren.

    Advanced tracing toolkits like SystemTap can help you analyze your program in real time, without modifying or restarting it. But they can also seem dauntingly unfriendly, especially when applied to interpreted languages like Python.

    Fear not! We'll talk about how kernel tracing actually works, what tools are available, and what we need to know about the Python interpreter's internals to use them effectively. We'll see how to do mixed-mode profiling, and how to trace specific events, like memory allocations or network calls. We'll discuss some of the pros and cons of these techniques, and how they can be applied to debugging systems in other languages too.

    ​

    • Meetup_square
    • Black Facebook Icon
    • Black Twitter Icon