Much That We Have Gotten Wrong About SRE

1. Developers wanted to ship their produce

To the other side

2. Production never matches the development environment. It resembles, but cannot match

So they deployed people on the other side

3. But this process was slow, they wanted to deploy faster

So they deployed Continuous-Deployment (CI/CD)

4. To improve reliability, we got SRE to do this

SREs’ first job was to hold this ship, but that’s all where they got stuck at

5. What Site Reliability Engineers should’ve built is

SREs should’ve been *engineering* and *observing* the bridge, but instead they became the bridge