|
安装Veritas Cluster Server 4.0
硬件简述:
1. Node1: Netra 20 (2 X UltraSPARC-III+, 2048M RAM, 72G*2 HardDisk) 2. Node2: Netra 20 (2 X UltraSPARC-III+, 2048M RAM, 72G*2 HardDisk) 3. Shared Storage: D1000 (36G*3 HardDisk) 一. 安装 操作系统 安装2/04版本(2004年2月份)的Solaris 8。在安装过程中需要选择英文作为主要语言。 二. 安装EIS-CD 安装EIS-CD 2/04版本,EIS-CD用于设置使用cluster的环境变量。 三. 安装patch 为了避免CPU虚高的问题,需要安装117000-05补丁。该补丁可以从SUN公司官方网站下载。下载该补丁以后解压将生成117000-05目录。使用如下命令安装patch: patchadd 117000-05 四. 安装共享磁盘 在本次环境中,我们使用SUND1000盘阵作为共享磁盘,是SCSI接口的盘阵,如果是光纤接口的盘阵需要另行设置。 1. 给盘阵加电 2. 将Node1和盘阵用SCSI连线连接 3. Node1加电 4. 使Node1进入ok模式(在console窗口监控Node1的启动,当出现启动信息时,迅速按下Ctrl+Break,即可进入ok模式) 5. {0} ok probe-scsi-all 6. {0} ok boot ?Cr 7. 重新启动以后进入操作系统,使用format命令确认盘阵已经被此系统加载 8. Node1断电,Node2加电,重复4-7步,确认盘阵也可以被此台机器加载 9. 为了使两台机器同时读取共享存储,需要修改其中一台机器的SCSI ID。给Node1加电,进入ok模式(此时的状态应该是Node1,Node2,存储均已加电,Node2目前使用format命令已经可以观察到盘阵被正常加载,Node1处于ok模式)。 10. 设置Node1的SCSI ID为5(默认为7) {0} ok setenv scsi-initiator-id 5 11. 通过第5步中的屏幕信息我们可以知道该SCSI设备的系统表示符为/pci@8,700000/scsi@2,1,这个标识符将在下一步使用 12. 设置Node1的SCSI ID为5。(注意第2行的输入中" scsi-initiator-id",双引号之后有一个空格!) {0} ok nvedit 0: probe-all 1: cd /pci@8,700000/scsi@2,1 2: 5 " scsi-initiator-id" integer-property 3: device-end 4: install-console 5: banner 13. 保存第12步的配置 {0} ok nvstore 14. 设置环境变量 {0} ok setenv use-nvramrc? true use-nvramrc? = true {0} ok setenv auto-boot? true auto-boot? = true 15. 重新启动Node1 {0} ok reset-all 16. 在Node1,Node2上均用format命令检查,确认都成功挂载了共享存储 五. 安装VCS 1. 设置rhosts文件,在两台机器上都要执行,以下以Node1为例。 在/目录下添加rhosts文件,使远程登陆生效,用以简便地在两台机器上同时安装VCS root@uulab-s22 # echo “+” > /.rhosts root@uulab-s22 # more /.rhosts + 2. 设置多个网卡IP 由于Cluster系统中的心跳网卡最好是一个独立的网络设备,所以我们在两台机器上设置另外一个网卡和IP地址,专门用于心跳设备。在两台机器上都要执行,以下以Node1为例。 root@uulab-s22 # echo "uulab-p22" > /etc/hostname.qfe0 root@uulab-s22 # touch /etc/notrouter root@uulab-s22 # vi /etc/hosts 添加如下行: 192.168.0.6 uulab-p22 192.168.0.8 uulab-p23 root@uulab-s22 # vi /etc/netmasks 添加如下行: 192.168.0.0 255.255.255.0 root@uulab-s22 # sync root@uulab-s22 # reboot 3. 开始安装 VCS 4.0的安装已经可以简化到只使用一个命令,将同时安装Veritas Volume Manager 4.0, Veritas File System 4.0,Veritas Cluster Server 4.0以及其他一些相关 软件。以下命令只需要在一个节点中执行即可。 root@uulab-s22 # cd /opt/sf_ha.4.0.sol/storage_foundation root@uulab-s22 # ./installsf 安装过程一路都是选择或者按照提示输入,总的来说比较简单,不再赘述。 六. 创建磁盘组 VCS安装完毕以后,将要求使用如下命令重新启动所有节点: shutdown -y -i6 -g0 重启完毕,开始创建磁盘组。以下命令只需要在一个节点中执行即可。 使用format命令确认我们需要添加到磁盘组中的共享磁盘为c4t0d0,c4t8d0,c4t9d0 root@uulab-s22 # format Searching for disks...done AVAILABLE DISK SELECTIONS: 0. c1t0d0 /pci@8,600000/SUNW,qlc@4/fp@0,0/ssd@w2100000c50569190,0 1. c1t1d0 /pci@8,600000/SUNW,qlc@4/fp@0,0/ssd@w2100000c5056c1a7,0 2. c4t0d0 /pci@8,700000/scsi@2,1/sd@0,0 3. c4t8d0 /pci@8,700000/scsi@2,1/sd@8,0 4. c4t9d0 /pci@8,700000/scsi@2,1/sd@9,0 Specify disk (enter its number): root@uulab-s22 # vxdisksetup -i c4t0d0 如果此命令执行以后报错如下,那么参看附录中的解决方法1。 VxVM vxdisksetup ERROR V-5-2-3535 c4t0d0s2: Invalid dmpnodename for disk device c4t0d0. 如果此命令执行以后报错如下,那么参看附录中的解决方法2。 VxVM vxdisksetup ERROR V-5-2-1813 c4t0d0: Disk is part of ipasdg disk group, use -f option to force setup. root@uulab-s22 # vxdisksetup -i c4t8d0 root@uulab-s22 # vxdisksetup -i c4t9d0 root@uulab-s22 # vxdg init hlrdg hlrdg-01=c4t0d0 root@uulab-s22 # vxdg -g hlrdg adddisk hlrdg-02=c4t8d0 root@uulab-s22 # vxdg -g hlrdg adddisk hlrdg-03=c4t9d0 七. 创建卷 root@uulab-s22 # vxassist -g hlrdg -b make oradata_vol 15g layout=nostripe,nolog nmirror=2 & root@uulab-s22 # vxassist -g hlrdg -b make oraredo_vol 5g layout=nostripe,nolog nmirror=2 & root@uulab-s22 # vxassist -g hlrdg -b make oraarch_vol 8g layout=nostripe,nolog nmirror=2 & root@uulab-s22 # vxassist -g hlrdg -b make hlr_vol 4g layout=nostripe,nolog nmirror=2 & 八. 使用VxFS创建文件系统 root@uulab-s22 # mkfs -F vxfs -o bsize=8192,largefiles /dev/vx/rdsk/hlrdg/oradata_vol root@uulab-s22 # mkfs -F vxfs -o bsize=8192,largefiles /dev/vx/rdsk/hlrdg/oraredo_vol root@uulab-s22 # mkfs -F vxfs -o bsize=8192,largefiles /dev/vx/rdsk/hlrdg/oraarch_vol root@uulab-s22 # mkfs -F vxfs -o bsize=8192,largefiles /dev/vx/rdsk/hlrdg/hlr_vol 九. 配置VCS 我们的环境是创建了4个文件系统,分别是oradata_vol, oraredo-vol, oraarch_vol, hlr_vol,文件系统可以被使用需要挂载(mount)到相应的目录中。先创建相应的目录,在两台机器上都要执行,以下以Node1为例。(假设系统中dba组和oracle用户都已经创建) root@uulab-s22 # mkdir -p /opt/oracle/data root@uulab-s22 # mkdir -p /opt/oracle/redo root@uulab-s22 # mkdir -p /opt/oracle/arch root@uulab-s22 # mkdir ?Cp /opt/hlr root@uulab-s22 # chown oracle:dba /dev/vx/rdsk/hlrdg/oradata_vol root@uulab-s22 # chown oracle:dba /dev/vx/rdsk/hlrdg/oraredo_vol root@uulab-s22 # chown oracle:dba /dev/vx/rdsk/hlrdg/oraarch_vol root@uulab-s22 # chown oracle:dba /opt/oracle/data root@uulab-s22 # chown oracle:dba /opt/oracle/redo root@uulab-s22 # chown oracle:dba /opt/oracle/arch 修改/etc/VRTSvcs/conf/config/main.cf文件,这是VCS的配置文件,该文件的修改可以使用命令行(比如hagrp,hares等命令)修改,也可以使用任何文本编辑器(比如vi)直接修改,此处我们选择使用vi进行修改。粗体字部分为需要新添加的行。 修改完毕以后该文件如下: include "types.cf" cluster vcs_hlr_cluster ( UserNames = { admin = hijBidIfjEjjHrjDig } ClusterAddress = "10.7.1.7" --此处为Cluster环境的虚拟IP Administrators = { admin } CounterInterval = 5 ) system uulab-s22 ( ) system uulab-s23 ( ) group ClusterService ( SystemList = { uulab-s22 = 0, uulab-s23 = 1 } UserStrGlobal = "LocalCluster@https://10.7.1.7:8443;LocalCluster@https://10.7.1.7:8443;" AutoStartList = { uulab-s22, uulab-s23 } OnlineRetryLimit = 3 OnlineRetryInterval = 120 ) DiskGroup hlrdg ( DiskGroup = hlrdg MonitorReservation = 1 ) IP webip ( Device = eri0 Address = "10.7.1.7" NetMask = "255.255.0.0" ) Mount arch_mnt ( MountPoint = "/opt/oracle/arch" BlockDevice = "/dev/vx/dsk/hlrdg/oraarch_vol" FSType = vxfs FsckOpt = "-y" ) Mount data_mnt ( MountPoint = "/opt/oracle/data" BlockDevice = "/dev/vx/dsk/hlrdg/oradata_vol" FSType = vxfs FsckOpt = "-y" ) Mount hlr_mnt ( MountPoint = "/opt/hlr" BlockDevice = "/dev/vx/dsk/hlrdg/hlr_vol" FSType = vxfs FsckOpt = "-y" ) Mount redo_mnt ( MountPoint = "/opt/oracle/redo" BlockDevice = "/dev/vx/dsk/hlrdg/oraredo_vol" FSType = vxfs FsckOpt = "-y" ) NIC csgnic ( Device = eri0 ) VRTSWebApp VCSweb ( Critical = 0 AppName = vcs InstallDir = "/opt/VRTSweb/VERITAS" TimeForOnline = 5 RestartLimit = 3 ) VCSweb requires webip arch_mnt requires hlr_mnt data_mnt requires redo_mnt hlr_mnt requires hlrdg redo_mnt requires arch_mnt webip requires csgnic 修改完毕,使用如下命令进行语法检查: root@uulab-s22 # hacf -verify /etc/VRTSvcs/conf/config 十. 测试VCS 配置完成以后,同时重新启动两台机器,启动完毕,使用如下命令检查VCS是否运行正常。 root@uulab-s22 # hares -display -group ClusterService 正常情况应该如下显示: #Resource Attribute System Value VCSweb Group global ClusterService VCSweb Type global VRTSWebApp VCSweb AutoStart global 1 VCSweb Critical global 0 VCSweb Enabled global 1 VCSweb LastOnline global uulab-s22 VCSweb MonitorOnly global 0 VCSweb ResourceOwner global unknown VCSweb TriggerEvent global 0 VCSweb ArgListValues uulab-s22 vcs /opt/VRTSweb/VERITAS 5 VCSweb ArgListValues uulab-s23 vcs /opt/VRTSweb/VERITAS 5 VCSweb ConfidenceLevel uulab-s22 100 VCSweb ConfidenceLevel uulab-s23 0 VCSweb Flags uulab-s22 VCSweb Flags uulab-s23 VCSweb IState uulab-s22 not waiting VCSweb IState uulab-s23 not waiting VCSweb Probed uulab-s22 1 VCSweb Probed uulab-s23 1 VCSweb Start uulab-s22 1 VCSweb Start uulab-s23 0 VCSweb State uulab-s22 ONLINE VCSweb State uulab-s23 OFFLINE VCSweb AppName global vcs VCSweb ComputeStats global 0 VCSweb InstallDir global /opt/VRTSweb/VERITAS VCSweb ResourceInfo global State Valid Msg TS VCSweb RestartLimit global 3 VCSweb TimeForOnline global 5 VCSweb MonitorTimeStats uulab-s22 Avg 0 TS VCSweb MonitorTimeStats uulab-s23 Avg 0 TS # ………… 也可以使用df命令检查是否所有需要挂载的文件系统都已正确挂载在Node1上,而在Node2上则无法看到。 做切换测试: root@uulab-s22 # hagrp -switch ClusterService -to uulab-s23 正常情况大概是5秒左右,所有的资源包括IP,文件系统加载点都会转移到Node2上。 十一. 附录 解决方法1: root@uulab-s22 # vxdiskadm Volume Manager Support Operations Menu: VolumeManager/Disk 1 Add or initialize one or more disks 2 Encapsulate one or more disks 3 Remove a disk 4 Remove a disk for replacement 5 Replace a failed or removed disk 6 Mirror volumes on a disk 7 Move volumes from a disk 8 Enable access to (import) a disk group 9 Remove access to (deport) a disk group 10 Enable (online) a disk device 11 Disable (offline) a disk device 12 Mark a disk as a spare for a disk group 13 Turn off the spare flag on a disk 14 Unrelocate subdisks back to a disk 15 Exclude a disk from hot-relocation use 16 Make a disk available for hot-relocation use 17 Prevent multipathing/Suppress devices from VxVM's view 18 Allow multipathing/Unsuppress devices from VxVM's view 19 List currently suppressed/non-multipathed devices 20 Change the disk naming scheme 21 Get the newly connected/zoned disks in VxVM view 22 Change/Display the default disk layouts 23 Mark a disk as allocator-reserved for a disk group 24 Turn off the allocator-reserved flag on a disk list List disk information ? Display help about menu ?? Display help about the menuing system q Exit from menus Select an operation to perform: 17 Exclude Devices Menu: VolumeManager/Disk/ExcludeDevices VxVM INFO V-5-2-1239 This operation might lead to some devices being suppressed from VxVM's view or prevent them from being multipathed by vxdmp (This operation can be reversed using the vxdiskadm command). Do you want to continue ? [y,n,q,?] (default: y) y Volume Manager Device Operations Menu: VolumeManager/Disk/ExcludeDevices 1 Suppress all paths through a controller from VxVM's view 2 Suppress a path from VxVM's view 3 Suppress disks from VxVM's view by specifying a VID:PID combination 4 Suppress all but one paths to a disk 5 Prevent multipathing of all disks on a controller by VxVM 6 Prevent multipathing of a disk by VxVM 7 Prevent multipathing of disks by specifying a VID:PID combination 8 List currently suppressed/non-multipathed devices ? Display help about menu ?? Display help about the menuing system q Exit from menus Select an operation to perform: 5 Exclude controllers from DMP Menu: VolumeManager/Disk/ExcludeDevices/CTLR-DMP Use this operation to exclude all disks on a controller from being multipathed by vxdmp. As a result of this operation, all disks having a path through the specified controller will be claimed in the OTHER_DISKS category and hence, not multipathed by vxdmp. This operation can be reversed using the vxdiskadm command. VxVM INFO V-5-2-1263 You can specify a controller name at the prompt. A controller name is of the form c#, example c3, c11 etc. Enter 'all' to exclude all paths on all the controllers on the host. To see the list of controllers on the system, type 'list'. Enter a controller name [,all,list,list-exclude,q,?] c4 VxVM INFO V-5-2-1129 All disks on the following enclosures will be excluded from DMP ( ie claimed in the OTHER_DISKS category and hence not multipathed by vxdmp) as a result of this operation : Disk OTHER_DISKS Continue operation? [y,n,q,?] (default: y) y Do you wish to exclude more controllers ? [y,n,q,?] (default: n) n Volume Manager Device Operations Menu: VolumeManager/Disk/ExcludeDevices 1 Suppress all paths through a controller from VxVM's view 2 Suppress a path from VxVM's view 3 Suppress disks from VxVM's view by specifying a VID:PID combination 4 Suppress all but one paths to a disk 5 Prevent multipathing of all disks on a controller by VxVM 6 Prevent multipathing of a disk by VxVM 7 Prevent multipathing of disks by specifying a VID:PID combination 8 List currently suppressed/non-multipathed devices ? Display help about menu ?? Display help about the menuing system q Exit from menus Select an operation to perform: q VxVM vxdiskadm NOTICE V-5-2-1187 Please wait while the device suppression/unsuppression operations take effect.
责编:
微信扫一扫实时了解行业动态
微信扫一扫分享本文给好友
著作权声明:kaiyun体育官方人口 文章著作权分属kaiyun体育官方人口 、网友和合作伙伴,部分非原创文章作者信息可能有所缺失,如需补充或修改请与我们联系,工作人员会在1个工作日内配合处理。
|
最新专题
|
|