Andor Molnar created HBASE-29958:
------------------------------------
Summary: Improve log messages
Key: HBASE-29958
URL: https://issues.apache.org/jira/browse/HBASE-29958
Project: HBase
Issue Type: Sub-task
Reporter: Andor Molnar
Assignee: Anuj Sharma
*First*
{noformat}
2026-03-04T17:26:04,184 INFO [main] hbase.TableName: Meta table suffix value:
2026-03-04T17:26:04,185 INFO [main] hbase.TableName: Meta table name:
hbase:meta{noformat}
These messages are redundant. Instead write the following:
{noformat}
2026-03-04T17:26:04,184 INFO [main] hbase.TableName: Read Replica Cluster
suffix value: <blank>{noformat}
*Second*
{noformat}
2026-03-04T17:26:05,304 WARN [master/hbase-docker:16000:becomeActiveMaster]
util.FSUtils: Active Cluster Suffix File
file:/data-store/hbase/active.cluster.suffix.id is empty
2026-03-04T17:26:05,305 INFO [master/hbase-docker:16000:becomeActiveMaster]
master.MasterFileSystem: This is the active cluster on this storage location,
File Suffix dcf9e176-8437-49a9-9e91-b221b94e8fe3: : Suffix
dcf9e176-8437-49a9-9e91-b221b94e8fe3: :{noformat}
I think we should also indicate that this is related to Read Replica feature.
Instead:
{noformat}
2026-03-04T17:26:05,304 WARN [master/hbase-docker:16000:becomeActiveMaster]
util.FSUtils: Read Replica Cluster id file
file:/data-store/hbase/active.cluster.suffix.id is present and contains cluster
id: dcf9e176-8437-49a9-9e91-b221b94e8fe3:<blank>
2026-03-04T17:26:05,305 INFO [master/hbase-docker:16000:becomeActiveMaster]
master.MasterFileSystem: Read Replica Cluster: this is the active cluster on
this storage location with cluster id:
dcf9e176-8437-49a9-9e91-b221b94e8fe3:<blank> {noformat}
*Third*
Following is the error message which is shown when the second cluster is trying
to run in R/W mode:
{noformat}
java.io.IOException: Cannot start master, because another cluster is running in
active (read-write) mode on this storage location. Active Cluster Id: {}
dcf9e176-8437-49a9-9e91-b221b94e8fe3: This cluster Id:
dcf9e176-8437-49a9-9e91-b221b94e8fe3{noformat}
It doesn't make any sense. Instead we should see:
{noformat}
java.io.IOException: Cannot start master, because another cluster is running in
active (read-write) mode on this storage location. Active cluster id:
dcf9e176-8437-49a9-9e91-b221b94e8fe3:<blank>, this cluster id:
dcf9e176-8437-49a9-9e91-b221b94e8fe3:<replica1>{noformat}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
