unhealthy nodes that require continous rerun of host upgrade wizard to become healthy and able to run services

Hi,

I am having the above mentioned problem, and not sure what I missing. I
have two CentOS one master and one normal host, and three Ubunto nodes. I
am not sure what is wrong that require this, one of the nodes took longer
than the rest to install then I aborted, but in the first rerun of the
upgrade it picked up and became like the rest.

Any hint how to fix this is appreciated,

thank you very much,