- Set up PostgreSQL cluster on Kubernetes in 5 minutes (29 Jun 2020)
Managing the data layer is non-trivial because containers need redundancy and high-availability in order to be reliable. Enter Helm, the Kubernetes package manager, to make the orchestration simpler.
- Use free tools to provision infrastructure-as-code (17 Mar 2020)
Avoid click-and-tick web interfaces that lose or obfuscate config. Instead write simple code to provision your environments using free and open-source software (FOSS).
- Put containers into production, in weeks not months (04 Mar 2020)
If you've got a beta app but no production environment to host it in, start small and iterate towards a full Cloud
- Create an Oracle Cloud VM (always free) instance using Terraform (12 Feb 2020)
Oracle Cloud is still something of the new kid on the block, but their always-free tier offers a great starting place for some automated cloud resources provisioned with Terraform
- Beware open ports in Oracle Cloud's Default Security Lists (12 Feb 2020)
Setting up a Network Security Group confers additional (OR) access, not restrictive access (AND)
- Puppetting machines from a puppetmaster (16 Nov 2019)
Here are some exercises to get you using Terraform and Puppet together
- Git for Version Controlled magic (07 Oct 2019)
There are many version control systems (VCS) but few have the reach or convenience of git
- Manage secrets in Git repos using Eyaml (10 Sep 2019)
Strong asymmetric encryption can keep secrets confidential even if the repository that stores them is open source
- Create a Puppetmaster using Puppet masterless (17 Aug 2019)
A simple exercise in bootstrapping the most complex config-managed server in the estate, the Puppetmaster.
- Use Puppet for templating Hiera data into configuration files (14 Aug 2019)
Maintaining infrastructure-as-code at scale requires a clear, logical and ruthlessly consistent approach to machine state. Hiera is the way.
- Multiple node definitions in puppet's site.pp (13 Aug 2019)
A single node defined in site.pp is how all vast server estates begin, but multiple node definitions allow us to build out while sharing common configuration.
- Import Hiera values into Puppet classes (13 Aug 2019)
When configuration code contains 'hard-coded' configuration values, it's much harder to maintain. Hiera allows us to break the latter out into something readable.
- String substitution in Puppet using variables (12 Aug 2019)
Variables allow us to start genericising Puppet configuration and allow us to keep things DRY (don't repeat yourself)
- Create resource groups in Microsoft Azure using Terraform (11 Aug 2019)
Resource groups hold everything together in Azure, like virtual containers for the resources we provision.
- Organise Puppet manifests using classes (10 Aug 2019)
Classes are to resources like folders are to your Physics revision notes.
- Getting started with Puppet, masterless (07 Aug 2019)
Getting going with Puppet can be as simple a few lines of infrastructure-as-code (IAC).
- Set up Terraform to provision in Microsoft Azure (05 Aug 2019)
Terraforming your Azure Cloud means building out a more predictable, reliable and extensible Cloud estate for the future
- Puppet for configuration-managed Windows desktops (02 Aug 2019)
Configuration management makes homogenising similar machines easier and there are some great free tools to get started
- GraphQL using AWS AppSync (31 Jul 2019)
An AppSync GraphQL example, automatically provisioned with Terraform
- Get started with remote provisioning using Terraform (30 Jul 2019)
Often local machines or firewalls make it impossible to run Terraform locally, but remote provisioning offers a convenient secure alternative
- Force your attackers to cross multiple security boundaries (06 Jul 2019)
While a single firewall should be enough, the number of successful attacks implies that it's not.
- Keep track of AWS costs in real-time (27 Jun 2019)
Don't wait to find out tomorrow what you could learn today, or for a big bill from your Cloud provider at the end of the month
- Transferring Terraform state in S3 over to another project (22 Jan 2019)
Terraform state files in S3 are the way forward, but when you need to split a project, managing those state files requires care
- Cloud strategy inception (15 Oct 2018)
Know where you're going, but then focus on the first step not the destination.
- Web service review and Cloud audit (Checklist)
Beyond the code, there's so much that needs to be thought about when launching and running a production-grade, scalable and extensible app or web service. Let's get specific!
- Storing Terraform state in S3 (01 Oct 2018)
Terraform is a great solo tool but, when you start working together as a team, keep the tfstate files in the Cloud (S3)
- How to go to the Cloud, well (01 Oct 2018)
Cloud might feel like a tick-box purchasing decision, but getting tangible business outcomes from a Cloud strategy is all about the people.
- Cloud framework (Roadmap)
Start simple, start with people, but aim for a holistic well-constructed Cloud.
- Spinning up AWS Lambda functions (Part 1) (22 Apr 2018)
Serverless infrastructure holds great promise but building something fast that runs fast requires planning.
- Cost of Dev-Ops Collaboration (08 Apr 2018)
Collaboration is brilliant - the only way - but managing the volatility, uncertainty, confusion and ambiguity associated with it is expensive.
- Getting Started in the DevOps space (01 Apr 2018)
The Development-Operations world is a wash with some fantastic tools, but sometimes getting started can be too great an impediment to realise the 'DevOps' efficiency promise.
- Scalability trumps speed (25 Mar 2018)
Serverless promises a world without low-level worries, but what's it like at scale?
- Help! We can't find the DevOps people. Do we already have the DevOps people? (19 Mar 2018)
With market demand for Site Reliability engineers, DevOps engineers and Service Automation engineers eclipsing supply, at what point does cross-training become the only way to keep pace with the ever-more-automated world?
- Kubernetes for a migratable future (18 Feb 2018)
Serverless function-as-a-service (like AWS Lambda) is magic, but it's hard to commit at this early stage.
- Doing it Smart (10 Feb 2018)
It's great fun doing it clever, but sometimes keeping it simple is harder.
Here are a collection of posts from our team on DevOps culture, tools and collaboration.