2020 – The year of IaC?


As we enter November, the penultimate month in a rather turbulent year, my thoughts normally wander on to what the next year will bring tech wise for me. Despite a punishing year for many businesses, the big tech companies (on the whole) have continued to grow, AWS posted revenues $11.6bn for Q3 2020, up by 29% year on year and Google Cloud made $3.44bn compared to $2.38bn for same quarter in 2019. But forget about the big boys, what’s more interesting is those tech companies on the way up, and theres a few!

So its fair to say, 2020 has been a bit of a crazy one for just about everyone (unless of course you live under a rock?!) with technology coming to the aide of billions as video calling allowed people to stay social, online deliveries went crazy, and Joe Wicks became a household name in the UK with his daily lockdown exercise livestreams.

Whilst it’s been great to see my 85 year old Gran using whatsapp video calling, this technology has been around for a while and isn’t exactly new – especially not for us techie folk (I really do sound old with some of my word choices!). What is interesting for me, has been the hard push for automation platforms and the monetising of them.

Terraform, Ansible, SaltStack, Chef, Puppet, the list goes on. But I doubt there is a reader of this blog who hasn’t heard of any one of these products, hell I bet 99%+ are using them in some way, shape or form in their business right now. But a lot of these products have something in common – they are community driven. So what does that matter I hear you cry? Well, is something is free, and is fully functional – its challenging to make a profit out of it.

A quick google has given me some figures and patterns which surprised me hugely.

  • Hashicorp
    • Estimated value: $5.1bn (March 2020)
  • Ansible
    • Bought by Red Hat (now IBM) back in 2015 for $150m
    • 40% increase in revenue for Q1 2020
  • SaltStack
    • VMware announce intent to purchase in Sept 2020
    • Revenues were estimated at around $16m pa
  • Chef
    • Bought by Progress in Sept 2020 for $220m
    • Revenues were estimated to be around $70m pa
  • Puppet
    • Estimated revenues of $89m pa
    • Latest round of funding raised additional $42m from the likes of Cisco and VMware

The obvious takeaway from the above is that the behemouths of the IT industry are buying up these platforms en masse. But the clear message is the sheer amount of profit that is being achieved, by products which without community development hold little, to no value. But maybe that is the point?

If we look at how these products are typically designed, you can see why they are so popular. Lets take Hashicorp’s Terraform product, written in Go – a language similar in syntax to C, and designed at Google. The beauty of it being used in such a popular language, is that the community driven modules (or providers as Terraform calls them) are easy enough to write and the Terraform HCL language simplifies it so that even people new to IaC can get going.

Ansible in contrast, is based on Python, again a very common programming language but uses YAML as the input – keeping the learning curve really flat for new users, whilst appealing to the HUGE python community for modules.

I think when it comes to automation it seems like its more important to ensure that its accessible to the community for writing the modules than it is to have a finish product – or that’s how it seems to me.

So lets (totally hypothetically) say that you want to get started with IaC – you do your research to see which platform most suits your end goal, is it building infrastructure, is it configuration management or even patch remediation? You decide upon your platform of choice, follow a few handy-dandy blog guides, YouTube tutorials, or even an online course such as Pluralsight or Udemy. Then you start writing your code, using those community driven modules and that free-to-use language.

Monetising these products to make significant profit, especially in a very competitive market is a difficult feat, but its clear that some of these businesses as succeeding on a huge scale. This is interesting because automation spans two distinctly different areas of IT. Developers traditionally use free-to-use languages as they are builders from code level, whether that be Javascript, Python or PHP whereas Infrastructure engineers are builders in their own right, but are more likely to use Enterprise-ready software, with a company-backed support and secure platform. The increase in use of the paid-for automation platforms says that the products are spanning both areas and appealing to a lot of businesses which is a very unique proposition.

Hashicorp are one of those businesses that seem to excelling in 2020, and are uniquely positioned in that they haven’t been bought out by one of the huge IT corporations, they are still going through rounds of funding ($175m in March 2020). Another of their individuality is that they don’t just offer a single product, to date they have 8 products each with a specific feature set, whether that be infrastructure, networking, security or applications. Their increasing valuation to $5.1bn up from $1.9bn in 2018 shows its success and there has been a lot of excitement around Boundary, their latest product which focuses on using Identity to remotely access resources, as opposed to Bastion hosts or VPNs – very on-point given the challenges of remote access that has been the year of 2020.

Leave a Reply