## The problem
Issue fluxcd/flux2#5446 (closed, 18 comments): xref: https://github.com/fluxcd/helm-controller/issues/1086 Both the Kustomization and HelmRelease APIs have fields for referencing ConfigMaps and Secrets containing values used for templating, i.e. that have a direct impact on the rendered and applied resources. Right now, users are defining very short reconciliation intervals for getting better responsiveness in this process, i.e. if the content of the referenced ConfigMap/Secret changes, their controllers quickly reconcile their objects updating resources as needed.
## The fix
[stefanprodan (maintainer)]: To make ConfigMap and Secret changes trigger reconciliation, use the Flux v2.6 RC images (helm-controller rc-93880f81, kustomize-controller rc-56d3ffb8), which make valuesFrom, substituteFrom, and decryption secretRef reactive; casibbald confirmed watch events fire. If pods still do not restart after the reconcile, that is normal Kubernetes behavior, not a Flux bug: annotate the pod template with a checksum of the secret/configmap data, or use env instead of envFrom.
## Notes
Thread: https://github.com/fluxcd/flux2/issues/5446. Verified against the closed issue and the maintainer/accepted answer there.