简体繁体 English

Solrcloud-部队领袖选举

[英]Solrcloud- Force Leader election

原文 2018-08-01 04:39:06 2 1 solr/ solrcloud

We have solr cluster and one collection with 5 nodes. 我们有Solr集群和一个包含5个节点的集合。 Below are the example. 以下是示例。

Collection Name: list_test 集合名称：list_test

Node1: 192.168.1.1 -- Leader 节点1：192.168.1.1-领导者
Node2: 192.168.1.2 -- Active 节点2：192.168.1.2-活动
Node3: 192.168.1.3 -- Active 节点3：192.168.1.3-活动
Node4: 192.168.1.4 -- Active 节点4：192.168.1.4-活动
Node5: 192.168.1.5 -- Active 节点5：192.168.1.5-活动
Node6: 192.168.1.6 -- Active 节点6：192.168.1.6-活动
In some case Node1(Leader) server is down. 在某些情况下，Node1（主机）服务器已关闭。 Any one of nodes will became as a leader. 任何一个节点都将成为领导者。 After fixing the node1 issue the server became up to active node. 解决了node1问题后，服务器达到了活动节点的状态。
Our scenario is after fixing the node1 issue, we need node1 will back to as a leader node. 我们的方案是在解决了node1问题之后，我们需要将node1重新作为领导节点。 Is it possible? 可能吗？

Note: Solr version 7.4.0 注意：Solr版本7.4.0

1 个解决方案

You might want to try the new APIs for the leader election. 您可能要尝试使用新的API进行领导者选举。 I would suggest to take necessary replications enabled on your core or taking necessary backups or process to regenerate indexes in case of a Disaster (ie Disaster Recovery). 我建议您在核心上启用必要的复制，或者在发生灾难（即灾难恢复）时采取必要的备份或过程来重新生成索引。

Balance Shard Unique 平衡碎片唯一

Re-Balance Leaders 重新平衡领导者

Force Leaders 部队领袖

This is an expert level command, and should be invoked only when regular leader election is not working. 这是专家级别的命令，仅当常规领导者选举不起作用时才应调用此命令。 This may potentially lead to loss of data in the event that the new leader doesn't have certain updates, possibly recent ones, which were acknowledged by the old leader before going down. 如果新领导者没有某些更新（可能是最近的更新），则有可能导致数据丢失，旧领导者在停机之前已经确认了这些更新。