3 ms·

We've been running a very large production PyPy deployment across pretty much all our Python apps for about... 4 years now. Saves us a ton of money for essentia

We've been running a very large production PyPy deployment across pretty much all our Python apps for about... 4 years now. Saves us a ton of money for essentially no real downside.

makmanalp 9y ago
Just out of curiousity, would you be willing to answer a few more questions? What has the memory tradeoff been like? What is the workload you're using it for?
- JulianWasTaken 9y ago
  Certainly! It's a bit hard to answer some of those questions because it's been so long since we've run CPython, and also because we've now got ~10 apps or so that run on PyPy. Initially memory tradeoff was definitely significant, somewhere around 40% or so -- it's going to vary across applications though certainly, and in a lot of cases I'm a bit happy our memory usage went up because it forces us more towards "nicer" architectures where data and logic are cleanly separated. Not that I mean to apologize too much for it, it's something certainly to watch, but for us on our most widely deployed low-latency, high-throughput app, we traded about 40% speedup for 40% RAM on an app that does very little true CPU-bound tasks (it's an s2s webapp where per-request we essentially are doing some JSON parsing, pulling some fields out, building some data structures, maybe calling a database or two, and assembling a response to serialize ~500 times/sec/CPU core). On more CPU-bound workflows, like one we have that essentially just computes set memberships at 100% resource usage all day long, we saw multiplicative increases, and I can't even mention how much exactly, because the speedup was so great that we couldn't run it in our data center because it started using up all our bandwidth, so I only have numbers for once it was moved into AWS and onto different machines :). Happy to elaborate more, as you can tell, I think companies with performance-sensitive workloads need to be looking at PyPy, so always happy to talk about our experiences.
  - sandGorgon 9y ago
    are you using any math packages like Numpy/Pandas or Opencv ?
    - JulianWasTaken 9y ago
      Not in any production workloads. They do work these days in PyPy though, so I'd feel comfortable doing so if we did, although I'd probably feel just as comfortable writing whatever numerics in pure-Python too unless it was stuff that already existed easily elsewhere. On a personal note I've played with OpenCV as well (and done so with PyPy to do some real-time facial analysis on a video stream), but yeah also not for $PRODUCTION_WORK.

We've been running a very large production PyPy deployment across pretty much all our Python apps for about... 4 years now. Saves us a ton of money for essentially no real downside.

makmanalp 9y ago

Just out of curiousity, would you be willing to answer a few more questions? What has the memory tradeoff been like? What is the workload you're using it for?

JulianWasTaken 9y ago
Certainly! It's a bit hard to answer some of those questions because it's been so long since we've run CPython, and also because we've now got ~10 apps or so that run on PyPy. Initially memory tradeoff was definitely significant, somewhere around 40% or so -- it's going to vary across applications though certainly, and in a lot of cases I'm a bit happy our memory usage went up because it forces us more towards "nicer" architectures where data and logic are cleanly separated. Not that I mean to apologize too much for it, it's something certainly to watch, but for us on our most widely deployed low-latency, high-throughput app, we traded about 40% speedup for 40% RAM on an app that does very little true CPU-bound tasks (it's an s2s webapp where per-request we essentially are doing some JSON parsing, pulling some fields out, building some data structures, maybe calling a database or two, and assembling a response to serialize ~500 times/sec/CPU core). On more CPU-bound workflows, like one we have that essentially just computes set memberships at 100% resource usage all day long, we saw multiplicative increases, and I can't even mention how much exactly, because the speedup was so great that we couldn't run it in our data center because it started using up all our bandwidth, so I only have numbers for once it was moved into AWS and onto different machines :). Happy to elaborate more, as you can tell, I think companies with performance-sensitive workloads need to be looking at PyPy, so always happy to talk about our experiences.
- sandGorgon 9y ago
  are you using any math packages like Numpy/Pandas or Opencv ?
  - JulianWasTaken 9y ago
    Not in any production workloads. They do work these days in PyPy though, so I'd feel comfortable doing so if we did, although I'd probably feel just as comfortable writing whatever numerics in pure-Python too unless it was stuff that already existed easily elsewhere. On a personal note I've played with OpenCV as well (and done so with PyPy to do some real-time facial analysis on a video stream), but yeah also not for $PRODUCTION_WORK.