# Unexpected cloud cost spike. How to handle postmortem? — retained record

> Retained reddit record from reddit:devops.

Canonical: https://highsignal.app/data/records/4619be71429d2776

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:devops)  published 2026-11-02 12:03 ingested 2026-11-03 02:38

[original source ↗](https://www.reddit.com/r/devops/comments/1xlvd19/unexpected_cloud_cost_spike_how_to_handle/)

## retained text



```
We had a recent incident where an auto-scaling config kept launching additional instances. No one noticed at first, and it happened long enough that our cloud services bill spiked pretty dramatically before somebody spotted the issue. I'm dreading having to explain this to higher ups, which is happening next week. We've already done a postmortem, but I want to be really prepared when I talk about this.

How have you handled similar situations? What did you change moving forward, especially related to identifying root causes and setting limits on resource scaling?
```



document fetched 2026-11-03 00:25 · canonical https://reddit.com/r/devops/comments/1xlvd19/unexpected_cloud_cost_spike_how_to_handle
