Blogging about your Cloud Tech is only interesting when it’s Novel

If you’re blogging about moving to the cloud, you have to write about the interesting things in your migration, and not just how you did Best-Practice.

So a while back I bitched about Why The Cloud Is Oversold, talking more generally about the supposed other-wordly experience that having Sensibly Flexible Virtualised IT is… well I’ve a new pet-hate: organisations Overselling Their Adoption Of The Cloud.

I know transparency is good. It’s also pragmatic because if the information is on a computer that is even near another computer that’s on the internet, it’s going to be leaked.1

It’s genuinely interesting when people share the unique work they’ve done. Especially when Public Bodies do stuff, look at how much gov.uk open-sourced, and how much of that govt.nz reused. We’ll not mention that Scottish Government developers can’t access the gov.uk repo as GitHub is a blocked “file-sharing” site.

The team I worked with at the BBC have spoken widely about how they turn ongoing streams of video into neatly segmented files, that are uploaded to S3 at more than 1 gigabit a second, and how these are made into the things you see on /iplayer.2

Alongside the stuff that’s of sufficient scale to be interesting, Video Factory also uses a load of standard enterprise patterns: micro-services, communication through queues, separation of concerns, etc… They’ve spoken about these, but very much in a “we’re just doing best-practice after big monolithic system pissed us off too much” way.

Anyway, I just read a blog post, by another public body documenting their transition to the Cloud and a new Responsive Website.3

Turns out sometimes they get a lot of load, and this is a problem they’ve had to solve. I’ll give you a second to think about how you’d solve bursty load on AWS.

Have you guessed?

They’ve only cached the site behind varnish, and are running that in an auto-scale group behind a Elastic Load Balancer.

That’s a pretty standard best-practice. Perhaps the novelty is that they’re a Public Sector body doing a sensible thing.4

But best-practice, by its very definition, just isn’t interesting blog-fodder: “Hey, We Do The Thing That Everyone Else Is Doing”.5

This leaves me wondering what next from this organisation:

  • “Our Windows PC Estate uses Microsoft Update Server to ensure they’re patched”
  • “We make our endpoints run anti-virus and disable USB ports on front-line single-use machines”
  • “We use Active-Directory federation to provide single-signon across all of our desktop applications”

If we’re really lucky maybe they’ll tell us: “How We Use Chaos-Monkey to Simulate Cloud Error-Situations”

I can’t wait.

  1. That is an exaggeration, but not nearly as much as I’d like it to be
  2. I helped make this bit and I’m still disproportionately proud of it
  3. The kind you hate on the desktop because of all the white-space, and where the custom fonts don’t look quite right
  4. I could link to numerous projects here, so here is a small selection of failure
  5. Netflix get to do it, because they’re the one of the groups setting out best-practice in AWS