Problem
During the adopt-a-cluster phase it checks your inventory.yaml
, cluster.yaml
and the runs
folder. It is possible that the operation would fail due to a corrupted runs folder.
An example of the error message shows:
error getting PreprovisionedInventory objects: %!w(<nil>)
Solution
Before anything else, check your cluster.yaml
file and inventory.yaml
for any stray/invalid characters or incorrect values. If everything checks out, you can take the following steps to "rebuild" your runs folder:
- Reverse or undo the changes you did to
cluster.yaml
on step 2 of https://archive-docs.d2iq.com/dkp/konvoy/2.1/major-version-upgrade/upgrade/preprovisioned/ to revert to your originalcluster.yaml
file. - Backup your
runs
foldermv runs runs_backup
- Re-run
konvoy up
to rebuild theruns
folder
Once done, re-run the dkp --kubeconfig=admin.conf prepare-to-adopt cluster preprovisioned
to see if you encounter further issues.