[P1BIGROCKS-1362/P1BIGROCKS-1686] Increased Auditing and Alerting
View options
Cluster Auditor was built to provide insights into non-compliance of clusters and should be expanded on
- Evaluation of Distro capabilities (RWX, Kubernetes versions, node health, etc.)
- Better validation of Iron Bank images. Look at Cosign to have standalone signed images. Notary only validates the pull, which can't be done when deploying with an Airgap Architecture. Look at including a manifest of images as part of the Big Bang deployment and validating against a finite list of SHAs
- Ensure email/mattermost/webhook alerting functions with Cluster Auditor events.
Although alertmanager is used as the alerting mechanism deployed within BB, it is rarely used outside of the defaults. This epic is to introduce sane rules for which to alert on by default.
Each package should have:
- sane defaults for alerting (if this means no alerting that is also okay)
Pre-requisites
This epic depends on &54 (closed) to be completed before beginning