feat(mgmt-agent): add node-health detection and labeling controller (AROSLSRE-1588) - #6284

Open

raelga wants to merge 18 commits into

raelga:raelga/aroslsre-1588-node-healer

Commits on Jul 31, 2026

refactor(mgmt-agent): move node-health detectors into their own package with a README
raelga
committed
refactor(mgmt-agent): rework node-health detection to pod-condition signals and Source.Host indexing (AROSLSRE-1588)
raelga
committed
refactor(mgmt-agent): record per-node success history for node-health detection (AROSLSRE-1588)
raelga
committed
test(mgmt-agent): add node-health controller tests (AROSLSRE-1588)
raelga
committed
fix(mgmt-agent): correlate node-health events by pod UID and require per-pod sustained dwell for the floor (AROSLSRE-1588)
raelga
committed
fix(detectors): do not count terminal-phase pods as stuck (AROSLSRE-1588)
raelga
committed
fix(controllers): retire stale wedged labels and dedup transitions (AROSLSRE-1588)
raelga
committed
docs(controllers): clarify configmap ownership and metric semantics (AROSLSRE-1588)
raelga
committed

Commits on Aug 1, 2026