From ad29dddc5f97cf6182e69c8f90317fddccca724b Mon Sep 17 00:00:00 2001 From: wu xiangkai Date: Fri, 17 Oct 2025 16:28:38 +0800 Subject: [PATCH] =?UTF-8?q?doc:=20=E9=98=85=E8=AF=BBmysql=20group=20replic?= =?UTF-8?q?ation=E6=96=87=E6=A1=A3?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- mysql/mysql集群/Mysql Group Replication.md | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/mysql/mysql集群/Mysql Group Replication.md b/mysql/mysql集群/Mysql Group Replication.md index 69c14a4..9e33b92 100644 --- a/mysql/mysql集群/Mysql Group Replication.md +++ b/mysql/mysql集群/Mysql Group Replication.md @@ -242,4 +242,8 @@ group members不仅需要对事务的提交达成一致,也需要对当前的v 通过`group_replication_member_expel_timeout`的system variable,引入了一个waiting period,允许members该时间内重新连接到group,从而避免被从group中驱逐。如果member在该时间范围内重新连接到了group,那么该member可以重新在group中处于活跃状态。但是,当member超过expel timeout并且从group中被驱逐时,或通过`STOP GROUP_REPLICATION`语句停止group replication时,或server failure时,其必须作为新的实例加入到group中。 ### Failure detection -group replication failure detection mechanism是一个分布式的service,用于检测group中server没有和其他servers沟通的情况,从而判定该server有停止服务的嫌疑。如果在group的共识中该嫌疑为true,那么group将会做出协调一致的决定,将该member逐出group。 \ No newline at end of file +group replication failure detection mechanism是一个分布式的service,用于检测group中server没有和其他servers沟通的情况,从而判定该server有停止服务的嫌疑。如果在group的共识中该嫌疑为true,那么group将会做出协调一致的决定,将该member逐出group。 + +将不参与communicating的member驱逐出group是必要的,因为group在对transaction或view change需要majority of members达成一致。如果member不参与这些决策,group必须对member进行移除,从而增加group中包含majority of correctly working members的可能性,从而能够继续处理transactions。 + +在replication group中,每两个member间都有一个point-to-point的communication channel,从而形成了一个full connected graph。这些连接通过group communication engine(XCom, a Paxos variant)来进行管理,来凝结使用TCP/IP sockets。 \ No newline at end of file