[
https://issues.apache.org/jira/browse/HBASE-19293?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16258717#comment-16258717
]
Sean Busbey commented on HBASE-19293:
-------------------------------------
-1 on v1 and v2 patches.
1) please do not remove default methods from IA.Public interfaces (i.e. Admin).
That will break folks who rely on that default being present.
2) the current QA flags are from the patch. The checkstyle one won't show up if
you back out the removal of defaults from Admin.
3) could we keep removal of deprecated methods for 3.0 in a different patch?
It's easier to review the changes for this patch if they're just doing what it
says in the subject.
4)
{code}
diff --git a/hbase-protocol-shaded/src/main/protobuf/Replication.proto
b/hbase-protocol-shaded/src/main/protobuf/Replication.proto
index 7e78144..88efa00 100644
--- a/hbase-protocol-shaded/src/main/protobuf/Replication.proto
+++ b/hbase-protocol-shaded/src/main/protobuf/Replication.proto
@@ -77,6 +77,7 @@ message ReplicationHLogPosition {
message AddReplicationPeerRequest {
required string peer_id = 1;
required ReplicationPeer peer_config = 2;
+ required ReplicationState peer_state = 3;
}
{code}
This will make it so that we break admin clients further. Can we make the
peer_state optional and default to ENABLED?
> Support add a disabled state replication peer directly
> ------------------------------------------------------
>
> Key: HBASE-19293
> URL: https://issues.apache.org/jira/browse/HBASE-19293
> Project: HBase
> Issue Type: Improvement
> Reporter: Guanghao Zhang
> Assignee: Guanghao Zhang
> Fix For: 2.0.0-beta-1
>
> Attachments: HBASE-19293.master.001.patch,
> HBASE-19293.master.002.patch
>
>
> Now when add a replication peer, the default state is enabled. If you want
> add a disabled replication peer, you need add a peer first, then disable it.
> It need two step to finish now.
> Use case for add a disabled replication peer. When user want sync data from a
> cluster A to a new peer cluster.
> 1. Add a disabled replication peer. And config the table to peer config.
> 2. Take a snapshot of table and export snapshot to peer cluster.
> 3. Restore snapshot in peer cluster.
> 4. Enable the peer and wait all stuck replication log replicated to peer
> cluster.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
