Chinaunix首页 | 论坛 | 博客
  • 博客访问: 1047167
  • 博文数量: 74
  • 博客积分: 10000
  • 博客等级: 上将
  • 技术积分: 2950
  • 用 户 组: 普通用户
  • 注册时间: 2007-05-15 00:17
文章分类

全部博文(74)

文章存档

2009年(1)

2008年(73)

我的朋友

分类: 服务器与存储

2008-04-22 15:52:02

A split-brain condition (when one node survives and the other must be
    shut down, which can cause data corruption) might occurs when a Solaris
    host fails to correctly register with VERITAS vxfendg disks. The
    workaround is to clear the keys manually using the vxfenclearpre
    command when you encounter a split-brain condition. This command will
    remove Small Computer System Interface 3 (SCSI-3) registrations and
    reservations on the coordinator disks, as well as on the data disks in
    all shared disk groups.
   
    NOTE: Clearing the keys manually does not always work; sometimes it is
    necessary to bring all nodes down.
   
    To clear keys manually, perform the following steps:
   
    a. Shut down all other systems in the cluster that have access to the
       shared storage. This prevents data corruption.
    b. Start the script:
    # cd /opt/VRTSvcs/vxfen/bin
    # ./vxfenclearpre
    c. Read the script’s introduction and warning. Then, you can choose to
       let the script run.

    Do you still want to continue: [y/n] (default : n)
    y

    Note: Informational messages resembling the following might appear on
    the console of one of the nodes in the cluster when a node is ejected
    from a disk or LUN:
   
    scsi: WARNING: /sbus@3,0/lpfs@0,0/sd@0,1(sd91):
    Error for Command: Error
    Level:Informational
    scsi: Requested Block: 0 Error Block 0
    scsi: Vendor: Serial Number:
    0400759B006E
    scsi: Sense Key: Unit Attention
    scsi: ASC: 0x2a (), ASCQ:
    0x4,FRU: 0x0

    You may ignore these informational messages.

    Cleaning up the coordinator disks...
    Cleaning up the data disks for all shared disk groups...
    Successfully removed SCSI-3 persistent registration and reservations from
    the coordinator disks as well as the shared data disks.
    Reboot the server to proceed with normal cluster startup...
    #

    d. Restart all systems in the cluster.
   
阅读(1870) | 评论(0) | 转发(0) |
0

上一篇:买了台电脑

下一篇:什么是celogin

给主人留下些什么吧!~~