Synchronizing Django model definitions
This is about a small problem we faced with the models used for customers in YPlan, now Time Out Checkout.
(All tags.)
This is about a small problem we faced with the models used for customers in YPlan, now Time Out Checkout.
My plugin pytest-randomly was recently moved into the pytest-dev organization on GitHub, making it a bit “more official” as a pytest plugin. Thanks to Bruno Oliveira for suggesting it, Florian Bruhin and Bruno for approving it on the pytest-dev mailing list, and Gordon Wrigley for helping with its development.
Yesterday evening I gave a quiz at the London Django Meetup Group for the second year running. Here it is so you can do it at home (no cheating!). Answers are at the bottom.
Django’s ORM is normally fast enough as-is, but if you’ve ever profiled a high traffic view with a fairly complicated query, you might have found that constructing QuerySet can take a noticeable portion of your request time. For example, I once found a query on the front page of the site I was working on that took 1ms to construct and 1ms for the database to answer. With a performance budget of 100ms, that was 1% gone on computing the exactly same SQL.
During PyCon UK I had the opportunity to work on open-sourcing our in-house Django performance testing tool, which has now been released as django-perf-rec. We created it over two years ago, and have been using and improving it since. It has been helping us to pre-emptively fix performance problems in our code, and now it can help you!
We recently upgraded our 160,000 lines of backend Python code from Python 2 to Python 3. We did with zero downtime and no major errors! Here’s how we did it, hopefully it will help anyone else still stuck on Python 2!
I recently released version 0.1.10 of my library django-mysql, for which the main new feature was a backend for Django’s cache framework called MySQLCache. This post covers some of the inspiration and improvements it has, as well as a basic benchmark against Django’s built-in DatabaseCache.
On Tuesday I gave another talk at the London Django Meetup Group, titled “Factory Boy Fun”, based upon my previous blog post of the same name. The blog post covers pretty much the same stuff, but if you want to flick through the slides quickly for an overview, they’re hosted on GitHub.
I’ve recently been working on improving the test suite at YPlan. The biggest change is moving towards dynamic fixtures for our Django models using “Factory Boy”. This library is essentially a tool that lets you define simple helper functions to generate random, sensible model instances quickly; by using them in tests you can avoid the static JSON fixture files that Django recommends you use in tests by default. Factories are also general purpose - they just generate data and use it to create a model - and so they can be re-used to fill your development database rather than dumping from production.
On Tuesday I gave a talk on Django at the London Django Meetup Group, titled “Django at Scale.” The slides are hosted on GitHub.
I was looking through the MySQL slow_log for YPlan and discovered that there were a lot of SELECT COUNT(*) queries going on, which take a long time because they require a full table scan. These were coming from the Django admin, which displays the total count on every page.