Don’t overlook the second part:
Get them to watch this, if it doesn’t convince them, nothing will 
1 Like
This is the wildest take on this I have seen. No matter if you have 40 years under your belt as a coder you will still push broken stuff to prod unless you have the worlds simplest app.
Lightning fast CI and good telemetry is the strongest foundation of a resilient app in production.
3 Likes
You are right. I should’ve add a disclaimer that this was not a general advice but specific for this project / company and I did assume some things based on the opening post. Also mixed CI and CD. CI should always be as fast as possible (and run as often als feasible).
For CD it depends and I would still advise the topic author to not focus on it’s speed. Why? because the team seems to be under pressure. They don’t get the time to fix the current issues so fall back to ad-hoc solutions. A ‘slow’ CD can protect the team from such rush as “just push it” may become costly (and management hates taking a costly risk). The costlier a mistake is, the more time is allowed to be allocated for stability (within reason).
Edit: now I think of it; this does require disallowing SSH-ing into the machine to ‘manually’ fix things.
