Previously, when no pod with the appropriate label was found in the method kubernetesExecOne, the process would exit. This prevented the caller from retrying. I changed this to return with an error rather than exiting. I check all of the callers, and they can all handle the error return correctly.
Changes in this pull request:
Signed-off-by: Hal Rosenberg hrosenbe@vmware.com