Until there is a patch release, Elastic has documented this workaround:
# First, force upgrade the CRD chart to move to the v1 CRDshelm upgrade elastic-operator elastic/eck-operator-crds -n elastic-system --force# Then upgrade the main chart as usualhelm upgrade elastic-operator elastic/eck-operator -n elastic-system
bhearnmarked the checklist item Image versions updated in chart/values.yaml as completed
marked the checklist item Image versions updated in chart/values.yaml as completed
bhearnmarked the checklist item Chart version bumped (match upstream chart, reset to bb.0 if the upstream version changes) as completed
marked the checklist item Chart version bumped (match upstream chart, reset to bb.0 if the upstream version changes) as completed
bhearnmarked the checklist item Changelog entry added, note the update to image, any additional updates as completed
marked the checklist item Changelog entry added, note the update to image, any additional updates as completed
bhearnmarked the checklist item Pipeline passing on package MR as completed
marked the checklist item Pipeline passing on package MR as completed
bhearnmarked the checklist item Clean install tested and validated working (see things to test) as completed
marked the checklist item Clean install tested and validated working (see things to test) as completed
This version or newer will pull in the CRD updates from upstream that are needed for k8s 1.22+ (the older versions were deprecated and removed in 1.22).
While the upstream issue (or as they say "non-issue") hasn't been resolved I pulled this into the next sprint so that we can take another look and maybe settle on a workaround to get us to 1.22.
Unblocking for additional investigation since we'd like to be able to support 1.22 and not seeing any desire upstream to change the way this upgrade works.