What do you do when you've just been paged? Here is my checklist and some practical how-to for addressing production incidents 📋 Step 1: assess customer impact

Ideally you have a "vitals" dashboard ready to go that can give you confidence about the customer impact at a glance. There are a couple outcomes for this

If you're curious about ✨continuous delivery✨ and want to know everything about how cloud deployments work, oh boy should you strap in and read on, I have just the thread for you 🌈☁️🦄 Let's look at how Spinnaker performs a simple blue/green deployment on AWS.

This is our starting state, 20 instances for an application called egg, in the TEST account, in the us-east-1 region.

I am going to stop using phrases like "I didn't get anything done today" in status channels and standups.

It usually means "I didn't commit code" and contributes to the macho idea that only code matters.

1/n Instead I will start outlining the positive non-code contributions I made. Examples in no particular order:

✍️ writing a proposal
👋 giving feedback on proposals / docs / priorities / designs
🤓 representing the team in meetings
📖 writing documentation, runbooks

Short personal post-mosterm of how a small code change I made in Spinnaker caused Netflix to run with ~10k extra AWS instances overnight. Most deploys at Netflix use the red/black strategy, which causes us to temporarily use twice the needed capacity as the old and new server groups run side by side.