After yesterdays smug declaration of our Cloud-based technical prowess - we were quickly humbled when Amazon started having "increased error rates".

JJ accused me of breaking the internets.

"I hope you're proud of yourself."

Maybe a little. To wield such power with a simple blog post is awe-inspiring. I jest.

It's worth noting Cloudsmith wasn't badly affected. But some of our technology partners were; degrading our status down to amber.

It was an interesting situation - (our documentation framework provider) was completely down due to Heroku experiencing massive failures. Our S3 buckets were fine (we are currently hosted in Europe*) but the only AWS service we use that was affected was SQS. It experienced interruption.

Lessons Learned:

  1. You can never had too much redundancy.
  2. Always read the UPS manual (no matter how simple it seems).
  3. Don't write smug blog posts without handling points 1 and 2.

That is all.

*We may be currently hosted in Europe but our ultra-fast CDN means downloads are from local endpoints. Yay!