It seems like if the catalogd pod is re-spun while there is an unpacked ClusterCatalog on the cluster, the new catalogd pod starts responding with 404 errors from catalog clients.
The only way to resolve the situation seems to be to delete the ClusterCatalogs and re-create them.
It seems like if the catalogd pod is re-spun while there is an unpacked ClusterCatalog on the cluster, the new catalogd pod starts responding with 404 errors from catalog clients.
The only way to resolve the situation seems to be to delete the ClusterCatalogs and re-create them.