OM will NPE and crash when `ozone.om.service.ids=id1,id2` is configured but `ozone.om.nodes.id1` doesn't exist; or `ozone.om.address.id1.omX` doesn't exist. Root cause:
`OzoneManager#loadOMHAConfigs()` didn't check the case where `found == 0`. This happens when local OM doesn't match any `ozone.om.address.idX.omX` in the config.
Due to the refactoring done in
HDDS-2162. This fix has been included in that commit. I will repurpose the jira to add some tests for the HA config.