Shuo Huang created HDDS-17511:
---------------------------------
Summary: SCM can reject decommission despite sufficient healthy
datanodes
Key: HDDS-17511
URL: https://issues.apache.org/jira/browse/HDDS-17511
Project: Apache Ozone
Issue Type: Bug
Reporter: Shuo Huang
Assignee: Shuo Huang
The decommission check subtracts all IN_SERVICE target datanodes from the
healthy datanode count, including STALE or DEAD targets that were never
included in that count.
For example, the cluster has three healthy datanodes and one DEAD datanode. The
DEAD datanode's container list includes a container configured to keep three
replicas. Decommissioning that datanode is rejected with "Insufficient nodes":
the check calculates 3 - 1 = 2, although all three healthy datanodes remain
available.
The maintenance check has the same problem.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
