Siyao Meng created HDDS-17163:
---------------------------------
Summary: Expose count of at-risk containers whose loss of one more
replica would lose data
Key: HDDS-17163
URL: https://issues.apache.org/jira/browse/HDDS-17163
Project: Apache Ozone
Issue Type: Improvement
Components: SCM, metrics
Reporter: Siyao Meng
ReplicationManager reports under replicated and missing containers, but there
is no way to tell how many containers are one replica (or one EC chunk) away
from data loss. An under replicated RATIS/THREE container with 2 replicas left
and one with 1 replica left are counted the same.
Add a ReplicationManager metric for containers at risk:
* RATIS: only one healthy replica remains.
* EC: exactly d healthy replicas remain (e.g. 3 for EC 3+2).
This is the most important number to watch during disk or node failures and
decommission.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
