drbd主备集群脑裂处理过程

NFS00与NFS01使用drbd+keepalved构建双机主备drbd集群
这天巡视时发现同步失败了

[root@CDM-NFS01 szdc]# drbdadm status
r1 role:Primary
  volume:1 disk:UpToDate
  CDM-NFS00 connection:StandAlone

[root@CDM-NFS00 szdc]# drbdadm status
r1 role:Secondary
  volume:1 disk:UpToDate
  CDM-NFS01 connection:StandAlone

两台机器都处在StandAlone状态，此时判断一下是否处在脑裂状态

[root@CDM-NFS01 szdc]# grep dropping /var/log/message 
Split-Brain detected but unresolved, dropping connection!

故障处理过程：

# NFS00是备机，r1是资源名称
[root@CDM-NFS00 szdc]# drbdadm  connect --discard-my-data r1 

# NFS01是主机，重连一下资源
[root@CDM-NFS01 szdc]# drbdadm connect r1

# 两台机都恢复了
[root@CDM-NFS01 szdc]# drbdadm status
r1 role:Primary
  volume:1 disk:UpToDate
  CDM-NFS00 role:Secondary
    volume:1 peer-disk:UpToDate

[root@CDM-NFS00 szdc]# drbdadm status
r1 role:Secondary
  volume:1 disk:UpToDate
  CDM-NFS01 role:Primary
    volume:1 peer-disk:UpToDate
···