Resolving intermittent Fedora DNF error "No such file or directory: '/var/lib/dnf/'"

For many of my Ansible playbooks and roles, I have CI tests which run over various distributions, including CentOS, Ubuntu, Debian, and Fedora. Many of my Docker Hub images for Ansible testing include systemd so I can test services that are installed inside. For the most part, systemd-related issues are rare, but it seems with Fedora and DNF, I often encounter random test failures which invariably have an error message like:

No such file or directory: '/var/lib/dnf/'

The full Ansible traceback is:

Be careful, Docker might be exposing ports to the world

Recently, I noticed logs for one of my web services had strange entries that looked like a bot trying to perform scripted attacks on an application endpoint. I was surprised, because all the endpoints that were exposed over the public Internet were protected by some form of authentication, or were locked down to specific IP addresses—or so I thought.

I had re-architected the service using Docker in the past year, and in the process of doing so, I changed the way the application ran—instead of having one server per process, I ran a group of processes on one server, and routed traffic to them using DNS names (one per process) and Nginx to proxy the traffic.

In this new setup, I built a custom firewall using iptables rules (since I had to control for a number of legacy services that I have yet to route through Docker—someday it will all be in Kubernetes), installed Docker, and set up a Docker Compose file (one per server) that ran all the processes in containers, using ports like 1234, 1235, etc.

The Docker Compose port declaration for each service looked like this:

Revisiting Docker for Mac's performance with NFS volumes

tl;dr: Docker's default bind mount performance for projects requiring lots of I/O on macOS is abysmal. It's acceptable (but still very slow) if you use the cached or delegated option. But it's actually fairly performant using the barely-documented NFS option!

July 2020 Update: Docker for Mac may soon offer built-in Mutagen sync via the :delegated sync option, and I did some benchmarking here. Hopefully that feature makes it to the standard Docker for Mac version soon.

Ever since Docker for Mac was released, shared volume performance has been a major pain point. It was painfully slow, and the community finally got a cached mode that offered a 20-30x speedup for common disk access patterns around 2017. Since then, the File system performance improvements issue has been a common place to gripe about the lack of improvements to the underlying osxfs filesystem.

Molecule fails on converge and says test instance was already 'created' and 'prepared'

I hit this problem every once in a while; basically, I run molecule test or molecule converge (in this case it was for a Kubernetes Operator I was building with Ansible), and it says the instance is already created/prepared—even though it is not—and then Molecule fails on the 'Gathering Facts' portion of the converge step:

Running Drupal in Kubernetes with Docker in production

Update: Since posting this, there have been some interesting new developments in this area, for example:

  • There is now a Drupal/Kubernetes SIG which meets every other Wednesday.
  • There are Kubernetes Drupal Operators which can manage Drupal instances in Kubernetes; I maintain the geerlingguy/drupal-operator but there are a couple others out there in development.

Since 2014, I've been working on various projects which containerized Drupal in a production environment. There have always been a few growing pains—there will for some time, as there are so few places actually using Docker or containers in a production environment (at least in a 'cloud native' way, without tons of volume mounts), though this is changing. It was slow at first, but it's becoming much more rapid.

It's not me, Google, it's you - from GA to Fathom

tl;dr: I'm now using Fathom for my personal website analytics, and it's easy to self-host and maintain, better for privacy, and can lead to better site performance.

Since the mid-2000s, right after it became available, I started using Google Analytics for almost every website I built (whether it be mine or someone else). It quickly became (and remains) the de-facto standard for website usage analytics and user tracking.

Google Analytics UI

Before that you basically had web page visit counters (some of them with slightly more advanced features ala W3Counter and Stat Counter), and then on the high end you had Urchin Web Analytics (which is what Google acquired and turned into a 'cloud' version, naming the new product Google Analytics and tying it deeply into the Google AdWords ecosystem).

Deploying an Acquia BLT Drupal 8 site to Kubernetes

BLT to Kubernetes

Wait... what? If you're reading the title of this post, and are familiar with Acquia BLT, you might be wondering:

  • Why are you using Acquia BLT with a project that's not running in Acquia Cloud?
  • You can deploy a project built with Acquia BLT to Kubernetes?
  • Don't you, like, have to use Docker instead of Drupal VM? And aren't you [Jeff Geerling] the maintainer of Drupal VM?

Well, the answers are pretty simple:

Hosted Apache Solr now supports Drupal Search API 8.x-2.x, Solr 7.x

Earlier this year, I completely revamped Hosted Apache Solr's architecture, making it more resilient, more scalable, and better able to support having different Solr versions and configurations per customer.

Today I'm happy to officially announce support for Solr 7.x (in addition to 4.x). This means that no matter what version of Drupal you're on (6, 7, or 8), and no matter what Solr module/version you use (Apache Solr Search or Search API Solr 1.x or 2.x branches), Hosted Apache Solr is optimized for your Drupal search!

Hosted Apache Solr - version selection

The best way to get the full PHP version string

Recently, to automate building, tagging, and pushing my geerlingguy/php-apache Docker Hub image (see this issue), I needed to find a way to reliably determine the PHP major.minor.release version string. You'd think this would be simple.

Well, using Docker, I would run the image and then try:

# php --version
PHP 7.3.0-1+0~20181206202713.23+stretch~1.gbp076afd (cli) (built: Dec  6 2018 20:27:14) ( NTS )
Copyright (c) 1997-2018 The PHP Group
Zend Engine v3.3.0-dev, Copyright (c) 1998-2018 Zend Technologies
    with Zend OPcache v7.3.0-1+0~20181206202713.23+stretch~1.gbp076afd, Copyright (c) 1999-2018, by Zend Technologies

That's great; it outputs the version right at the start. But there are a few problems here:

Deploying a React single-page web app to Kubernetes

React seems to have taken the front-end development community by storm, and is extremely popular for web UIs.

It's development model is a breath of fresh air compared to many other tools: you just clone your app, and as long as you have Node.js installed in your environment, to start developing you run (either with npm or yarn or whatever today's most popular package manager is):

yarn install
yarn serve

And then you have a local development server running your code, which updates in real time when you change code.

But when it comes time to deploy a real-world React app to non-local environments, things can get a little... weird.

For most modern projects I work on, there are usually multiple environments: