How to change the Cluster Name in a Data Center deploy
プラットフォームについて: Data Center のみ - この記事は、Data Center プラットフォームのアトラシアン製品にのみ適用されます。
この KB は Data Center バージョンの製品用に作成されています。Data Center 固有ではない機能の Data Center KB は、製品のサーバー バージョンでも動作する可能性はありますが、テストは行われていません。サーバー*製品のサポートは 2024 年 2 月 15 日に終了しました。サーバー製品を利用している場合は、アトラシアンのサーバー製品のサポート終了のお知らせページにて移行オプションをご確認ください。
*Fisheye および Crucible は除く
目的
This guide will help you with the proper steps to update the cluster name of your cluster party.
ソリューション
First, we will need to update the confluence.cluster.name property of each Node to the name you want to use. This property is located in the <confluence-home>/confluence.cfg.xml file:
<property name="confluence.cluster.name">atlassian_cluster</property>- After updating the configuration file of each node to change the cluster name, please make sure to shut down the entire cluster (node by node)
Start each node at a time again, making sure not to start all nodes at the same time.
(Refer : How to : Stop/Start Confluence Application)
Note that if you restart a single node only or try to do a rolling restart after changing the cluster name, a cluster panic will be triggered and you will see the following log entries:
This log is from the node that got the cluster name updated and was restarted:
2019-02-13 14:20:56,505 INFO [localhost-startStop-1] [confluence.cluster.hazelcast.HazelcastClusterManager] configure Configuring Hazelcast with instanceName [confluence], join configuration TCP/IP member addresses: 192.168.56.100|192.168.56.101, network interfaces [fe80:0:0:0:65dc:2496:ae5d:8be%enp0s8, 192.168.56.100] and local port 5801 2019-02-13 14:20:56,505 INFO [localhost-startStop-1] [confluence.cluster.hazelcast.HazelcastClusterManager] startCluster Starting the cluster. 2019-02-13 14:20:59,115 WARN [hz.confluence.priority-generic-operation.thread-0] [com.hazelcast.cluster] log [192.168.56.100]:5801 [ciastes] [3.8.6] Node could not join cluster at node: [192.168.56.101]:5801 Cause: the target cluster has a different group-nameMeanwhile, this is the log you will see in the node that did not get renamed nor restarted, which is followed by the cluster panic error after a minute or so:
2019-02-13 14:20:59,037 WARN [hz.confluence.generic-operation.thread-0] [internal.cluster.impl.ClusterJoinManager] log [192.168.56.101]:5801 [ubuntu-cluster] [3.8.6] Received an invalid join request from [192.168.56.100]:5801, cause: clusters part of different cluster-groups 2019-02-13 14:20:59,038 WARN [hz.confluence.generic-operation.thread-1] [internal.cluster.impl.ClusterJoinManager] log [192.168.56.101]:5801 [ubuntu-cluster] [3.8.6] Received an invalid join request from [192.168.56.100]:5801, cause: clusters part of different cluster-groups 2019-02-13 14:20:59,044 WARN [hz.confluence.IO.thread-in-1] [hazelcast.nio.tcp.TcpIpConnection] log [192.168.56.101]:5801 [ubuntu-cluster] [3.8.6] Connection[id=2, /192.168.56.101:5801->/192.168.56.100:36469, endpoint=[192.168.56.100]:5801, alive=false, type=MEMBER] closed. Reason: Exception in NonBlockingSocketReader java.io.IOException: Connection reset by peer