Created on 02-20-2017 04:59 PM - edited 08-19-2019 04:01 AM
I've performed an upgrade of a cluster from HDP 2.4.2.0-258 to HDP 2.5.3.0-37. But for some reason, it triggered another Upgrade process. Everything upgraded, but now the cluster wants to perform another upgrade process.
cluster_version table
repo_version table
As you can see, upgrade 51 and 52 are the same. 51 completed.
Is there a way to mark the upgrade 52 as done or cancel/remove it?
Created 02-20-2017 05:12 PM
Can you do "select * from upgrade" table and check if any data exist in that table?
Also try executing below command -
--select distinct(request_id,status) from host_role_command where status IN ('PENDING', 'HOLDING_FAILED','ABORTED'); --update host_role_command set status = 'ABORTED' where status IN ('PENDING', 'HOLDING_FAILED') and request_id = 455;
Created on 02-20-2017 08:09 PM - edited 08-19-2019 04:00 AM
SELECT distinct(request_id,status) FROM host_role_command WHERE status IN ('PENDING', 'HOLDING_FAILED','ABORTED')
Created 03-06-2017 11:16 AM