模拟集群网络故障 - SAP HANA 开启 Amazon
Amazon Web Services 文档中描述的 Amazon Web Services 服务或功能可能因区域而异。要查看适用于中国区域的差异，请参阅 中国的 Amazon Web Services 服务入门 (PDF)。

本文属于机器翻译版本。若本译文内容与英语原文存在差异，则一律以英文原文为准。

模拟集群网络故障

描述 —模拟网络故障，以测试大脑分裂时的集群行为。

运行节点：可以在任何节点上运行。在这个测试用例中，这是在节点 B 上完成的。

运行步骤：

  • 使用以下命令丢弃所有来自和流向节点 A 的流量：

    iptables -A INPUT -s <<Primary IP address of Node A>> -j DROP; 
iptables -A OUTPUT -d <<Primary IP address of Node A>> -j DROP 
    [root@sechana ~]# pcs status
Cluster name: rhelhanaha
Stack: corosync
Current DC: prihana(version 1.1.19-8.el7_6.5-c3c624ea3d) - partition with quorum
Last updated: Fri Jan 22 14:45:24 2021
Last change: Fri Jan 22 14:45:11 2021 by hacluster via crmd on  sechana
2 nodes configured
6 resources configured
Online: [ prihana sechana ]
Full list of resources:
 clusterfence   (stonith:fence_aws):    Started prihana
 Clone Set: SAPHanaTopology_DRL_00-clone [SAPHanaTopology_DRL_00]
     Started: [ prihana sechana ]
 Master/Slave Set: SAPHana_DRL_00-master [SAPHana_DRL_00]
     Masters: [ prihana]
     Slaves: [ sechana ]
 hana-oip       (ocf::heartbeat:aws-vpc-move-ip):       Started prihana
Daemon Status:
  corosync: active/enabled
  pacemaker: active/enabled
  pcsd: active/enabled
[root@ sechana ~]#sechana:~ # iptables -A INPUT -s xxx.xxx.xxx.xxx -j DROP; 
iptables -A OUTPUT -d xxx.xxx.xxx.xxx -j DROP

预期结果：

  • 集群检测到网络故障并屏蔽节点 1。集群将辅助 SAP HANA 数据库（位于节点 2 上）提升为主数据库，使其接管为主数据库，而不会出现大脑分裂的情况。

    [root@sechana ~]# pcs status
Cluster name: rhelhanaha
Stack: corosync
Current DC: sechana (version 1.1.19-8.el7_6.5-c3c624ea3d) - partition with quorum
Last updated: Fri Jan 22 15:11:43 2021
Last change: Fri Jan 22 15:10:48 2021 by root via crm_attribute on sechana
2 nodes configured
6 resources configured
Online: [ sechana ]
OFFLINE: [ prihana]
Full list of resources:
 clusterfence   (stonith:fence_aws):    Started sechana
 Clone Set: SAPHanaTopology_DRL_00-clone [SAPHanaTopology_DRL_00]
     Started: [ sechana ]
     Stopped: [ prihana]
 Master/Slave Set: SAPHana_DRL_00-master [SAPHana_DRL_00]
     Masters: [ sechana ]
     Stopped: [ prihana]
 hana-oip       (ocf::heartbeat:aws-vpc-move-ip):       Started sechana
Failed Actions:
* clusterfence_monitor_60000 on sechana 'unknown error' (1): call=-1, 
status=Timed Out, exitreason='',
    last-rc-change='Fri Jan 22 14:59:14 2021', queued=0ms, exec=0ms
Daemon Status:
  corosync: active/enabled
  pacemaker: active/enabled
  pcsd: active/enabled
[root@sechana ~]#

恢复程序：

  • 清理集群 “failed actions”。