Why Your Kubernetes Automation Keeps Getting Blocked by Humans
A recent survey hit on something I see all the time in cluster work: teams know they need automation, but 71% still require human review before any resource changes get applied. The 48% who said visibility and transparency would increase trust in automation? That's the actual problem. When you're asking someone to approve a change they can't verify, you're not going to get buy-in. This isn't…
A recent survey revealed that 71% of teams still require human review before applying any changes to Kubernetes resources, despite knowing they need automation. The issue stems from a lack of visibility and transparency, as people are hesitant to approve changes they cannot verify. This isn't about trusting the automation algorithm, but rather about trusting the person asking them to approve blindly.
When a team writes a script to right-size resource requests based on actual usage, it can get stuck in approval limbo because reviewers don't understand the automation's reasoning. For instance, the reviewer sees a resource request change from 500m CPU to 200m, but the automation sees historical usage data of 180m CPU with room for bursts, indicating a recommendation of 250m CPU for headroom. Without visibility into the reasoning, approval becomes impossible.
To improve automation adoption, the survey suggests several practices:
1. Output the analysis, not just the result. Instead of simply generating a new resource request, provide a detailed analysis of the current state and the recommendation. For example, include CPU and memory usage patterns (P50, P95, P99), rationale for the recommendation, and the chosen value.
2. Use diff-friendly output to format recommendations as patches or diffs that are easy to review in Git. This allows reviewers to quickly see the differences between the current state and the proposed changes.
3. Implement a dry-run option that runs the automation by default but does not apply changes automatically. This allows team members to review the proposed changes before deciding whether to approve them. Even a simple --auto-approve flag, after a week of reviewing the automation's behavior, can greatly improve adoption.
The ultimate takeaway is that the automation gap exists because we treat "more automation" as the goal. However, if team members cannot see what the automation is doing, they will default to manual review every time. The solution isn't better algorithms, but rather better output that provides people with the necessary data and allows them to make informed decisions. This approach doesn't reduce automation; it makes automation more effective and widely adopted.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
