Engineering at Polyconseil

A technical insight into our staple food
  1. Migrating from json to jsonb

    How we migrated our PostgreSQL-based applications from text-formatted json to jsonb.

  2. Migrating from Ansible Tower to AWX

    For several years now, we used Ansible Tower to launch and manage our Ansible jobs. The GUI wasn't good and the software wasn't perfect, but it had some very important features for us. When RedHat bought Ansible, they multiplied our licence fee by 3. We told them everything that was wrong with Ansible Tower and paid for one more year. As we did not observed any improvement regarding our complains, we started to look elsewhere. When Redhat released an open source project around Ansible Tower codebase, we jumped in.

  3. SMTP as an API

    Using SMTP as a base protocol for machine or human API.

  4. Handling locks during database schema updates

    Database schema updates are inevitable. And we need to perform these updates while keeping the system up and responding. This can be a little scary at times, particularly when you remember this cloudy Tuesday when dozens of your customers called because everything seemed broken after running such a migration (I know you've been there too).

  5. Cross compiling with Schroot

    When you're dealing with embedded systems you will most likely need to cross compile software. Cross compiling and building on any computer or server can be a pain and the use of schroot is a good way to avoid them. We'll show a few examples of cross compiling software and finally we'll review how to do it with schroot and talk about pros and cons of this technique.

  6. Serious i18n text search

    Use Django, nece and PostgreSQL for serious internationalized text search and completion.

  7. Code your JS app like it's 86

    How to avoid rewriting your JavaScript application, thanks to older coding paradigms.

  8. Introducing sphinx-dust

    We wrote sphinx-dust, a small Sphinx extension that encourages developers to frequently review their documentation and keep it up-to-date. CI builds can then be configured to fail and notify use when a page becomes outdated.

  9. PostgreSQL fine-tuning

    A survey of our strategy in failure and disaster recovery, as well as the configuration of our PostgreSQL instances. Our goal is to achieve a high level of reliability while letting PostgreSQL operate at maximum speed, mostly in-memory, running replication and on-disk persistence asynchronously.

  10. Filtering Docker communications

    How we implemented network filters on our integration testing virtual machine, to avoid accidentally hitting production services or paying APIs. This article presents our iptables and privoxy setups.

Page 1 / 2 »