[Ajs1]

Hi, these 2 blogs list some scenarios. The first has more details because we replicated these in house and can share full details: https://www.zebrium.com/blog/is-autonomous-monitoring-the-an.... https://www.zebrium.com/blog/beyond-anomaly-detection-how-in...

[Ajs1]

and although this isn't quite answering your question about unknown/unknowns, this open source project lets you inject failure modes using a chaos tool (litmus) on your own app. We had really good results catching application incidents created by these chaos test. https://github.com/zebrium/zebrium-kubernetes-demo