[
https://issues.apache.org/jira/browse/HBASE-26481?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17467435#comment-17467435
]
Hudson commented on HBASE-26481:
--------------------------------
Results for branch master
[build #483 on
builds.a.o|https://ci-hadoop.apache.org/job/HBase/job/HBase%20Nightly/job/master/483/]:
(/) *{color:green}+1 overall{color}*
----
details (if available):
(/) {color:green}+1 general checks{color}
-- For more information [see general
report|https://ci-hadoop.apache.org/job/HBase/job/HBase%20Nightly/job/master/483/General_20Nightly_20Build_20Report/]
(/) {color:green}+1 jdk8 hadoop3 checks{color}
-- For more information [see jdk8 (hadoop3)
report|https://ci-hadoop.apache.org/job/HBase/job/HBase%20Nightly/job/master/483/JDK8_20Nightly_20Build_20Report_20_28Hadoop3_29/]
(/) {color:green}+1 jdk11 hadoop3 checks{color}
-- For more information [see jdk11
report|https://ci-hadoop.apache.org/job/HBase/job/HBase%20Nightly/job/master/483/JDK11_20Nightly_20Build_20Report_20_28Hadoop3_29/]
(/) {color:green}+1 source release artifact{color}
-- See build output for details.
(/) {color:green}+1 client integration test{color}
> Consider rolling upgrading from old region replication framework
> ----------------------------------------------------------------
>
> Key: HBASE-26481
> URL: https://issues.apache.org/jira/browse/HBASE-26481
> Project: HBase
> Issue Type: Sub-task
> Components: read replicas
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Priority: Major
> Fix For: 3.0.0-alpha-3
>
>
> In the old implementation, we will cerate a replication peer for region
> replication. And in the new framework, we do not need the peer any more, so
> we have deleted the RegionReplicaReplicationEndpoint class.
> But when rolling upgrading, the peer is still there so we will try to load it
> and it will lead to a class not found exception. We need to handle it.
> And during the upgrading, some of the region servers may not have the correct
> replicateToReplica method yet. Maybe we could try to catch the no such method
> error and try to fallback to use replay method. Although it is not perfect,
> but finally a flush can fix everything.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
