Veritas Volume Replicator VVR quick start

=== setup vvr
/etc/init.d/vras-vradmind.sh start                                              [master/secondary]
/etc/init.d/vxnm-vxnetd start                                                   [master/secondary]
vxassist -g repdg make mysrl 200M alloc=disk02                                  [master/secondary]
vradmin -g repdg createpri myrvg vol01 mysrl                                    [master]
vxdg list

put dgid into /etc/vx/vras/.rdg on the secondary site. eg
echo “1169092729.31.slave” >> /etc/vx/vras/.rdg  (on secondary only)            [secondary]
vradmin -g repdg addsec myrvg slave test01 prlink=link_slave srlink=link_test01 [master]
vradmin -g repdg -a startrep myrvg test01                                       [master]
===checking

vradmin -g repdg -l printrvg myrvg
vradmin -g repdg repstatus myrvg
vxrvg -g repdg getdatavols myrvg
vxrvg -g repdg getrlinks myrvg
vxrlink -g repdg status link_slave

vradmin -g repdg pauserep myrvg test01                                          [master]
vradmin -g repdg resumerep myrvg test01                                         [master]

===remove vvr

vradmin -g repdg -f stoprep myrvg test01                                        [master]
vradmin -g repdg delsec myrvg test01                                            [master]
vradmin -g repdg -f delpri myrvg                                                [master]

vxedit -g repdg -f -r rm mysrl                                                  [master/secondary]

vxassist -g repdg remove log vol01  [delete one log each time],  need to run twice if there are 2 logs   [master/secondary]

Veritas Cluster Server tip to manually import type/agent

It’s convenient to import a new type from hagui, but under some case you might be unable to do so, like no Xserver, or slow network access etc.

Here I will show you how to import  the new agent type from command line.

/opt/VRTSvcs/bin/hacf -typetocmd /etc/VRTSvcs/conf/VVRTypes.cf -display | ksh

easy and straightforward.  but maybe  you don’t know it.

hacf -typetocmd is not showing from “hacf -help”.

test01:/opt/VRTSvcs/bin #hacf -help
VCS INFO V-16-1-10601 Usage:
         hacf -cftocmd   <conf_dir> [-dest <dest_dir>] [-display]
         hacf -cmdtocf   <conf_dir> [-dest <dest_dir>]
         hacf -verify    <conf_dir> [-display]
         hacf -help

test01:/opt/VRTSvcs/bin #

vxvm snapshot quick start

source dg: testdg
source vol:  vol01
data volmune mounted at:  /slavefs
snapshot mounted at:  /snapfs

 

space-optimized instant snapshot
====================================
1. check dco
vxprint -g testdg -F%instant vol01
vxprint -g testdg -F%fastresync vol01
vxprint

2.if dco not available, create it by prepare
vxsnap -g testdg prepare vol01
vxprint -g testdg -F%instant vol01
vxprint -g testdg -F%fastresync vol01

3. create cache volume in preparation of space-optimized snap
vxassist -g testdg make cachevol 100M layout=nostripe init=active disk01
normally it should be 10% of original volume size

4. create cache objects in preparation of space-optimized snap
vxmake -g testdg cache mycacheobj cachevolname=cachevol autogrow=on
vxcache -g testdg start mycacheobj
vxprint -g testdg

5. create the instant snap with space-optimized
vxsnap -g testdg make source=vol01/newvol=snapvol01/cache=mycacheobj alloc=disk02
vxprint -g testdg -F%incomplete snapvol01
6. check file system
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01

7. mount file system
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs/

8. refresh
umount /snapfs
vxsnap -g testdg refresh snapvol01 source=vol01
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs/

9. restore data volume
umount /slavefs  (data volume)
vxsnap -g testdg restore vol01 source=snapvol01
fsck -F vxfs /dev/vx/rdsk/testdg/vol01
mount -F vxfs /dev/vx/dsk/testdg/vol01 /slavefs

10. remove instant space-optimized snapshot
umount /snapfs
vxedit -g testdg -r -f rm snapvol01     [-f is used as it is still enabled, otherwise stop it first]
vxedit -g testdg -f -r rm mycacheobj
vxsnap -g testdg -f unprepare vol01

 

 
full-sized instant snapshot
====================================
source dg: testdg
source vol:  vol01
1. check dco
vxprint -g testdg -F%instant vol01
vxprint -g testdg -F%fastresync vol01
vxprint

2.if dco not available, create it by prepare command
vxsnap -g testdg prepare vol01
vxprint -g testdg -F%instant vol01
vxprint -g testdg -F%fastresync vol01

3. create a full-sized volume in preparation of space-optimized snap
LEN=`vxprint -g testdg -F%len vol01`
DCONAME=`vxprint -g testdg -F%dco_name vol01`
RSZ=`vxprint -g testdg -F%regionsz $DCONAME`
vxassist -g testdg make snapvol01 $LEN layout=nostripe init=active alloc=disk02
vxsnap -g testdg prepare snapvol01 regionsize=$RSZ
4. create the snapshot
vxsnap -g testdg make source=vol01/snapvol=snapvol01 [syncing=off]
vxprint -g testdg -F%incomplete snapvol01

5. check file system
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01

6. mount file system
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs

7. refresh
umount /snapfs
vxsnap -g testdg refresh snapvol01 source=vol01
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs/
8. restore data from snap
umount /slavefs 
vxsnap -g testdg restore vol01 source=snapvol01
fsck -F vxfs /dev/vx/rdsk/testdg/vol01
mount -F vxfs /dev/vx/dsk/testdg/vol01 /slavefs
9. remove snap
vxprint -g testdg
umount /snapfs
vxedit -g testdg -f -r rm snapvol01
vxsnap -g testdg -f unprepare vol01

 

third-mirror break-off snapshots (must be instant ready)
=================================
1. dco/drl
vxsnap -g testdg prepare vol01
2. create mirror and run it in background
vxsnap -b -g testdg addmir vol01 nmirror=1 alloc=disk01,disk02  
You must spread into two disk for mirrored volume, as dcl is mirrored.

or

1. prepare volume
vxsnap -g testdg prepare vol01 ndcomirs=1 alloc=disk01

2. create mirror
vxsnap -b -g testdg addmir vol01 nmirror=1 alloc=disk02

 

continue
3. vxprint -g testdg  [to find out the plex with snapdone]/ vxtask list

4. vxsnap -g testdg make source=vol01/newvol=snapvol01/plex=vol01-02

5. fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01

6. mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs

7. refresh
umount /snapfs
vxsnap -g testdg refresh snapvol01 source=vol01
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs/

8. restore
umount /slavefs 
vxsnap -g testdg restore vol01 source=snapvol01
fsck -F vxfs /dev/vx/rdsk/testdg/vol01
mount -F vxfs /dev/vx/dsk/testdg/vol01 /slavefs
9. remove snap
vxprint -g testdg
umount /snapfs
vxedit -g testdg -f -r rm snapvol01
vxsnap -g testdg -f unprepare vol01

 
traditional third-mirror break-off snaphost
=============================================

1
vxassist -b -g testdg snapstart nmirror=1 vol01

2.
vxassist -g testdg snapshot vol01 snapvol01

3.
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01

4.
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs

5. refresh
umount /snapfs
vxassist -b -g testdg snapback snapvol01
vxassist -g testdg snapshot vol01 snapvol01
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs

6. restore
umount /slavefs
umount /snapfs
vxassist -b -g testdg -o resyncfromreplica snapback snapvol01
fsck -F vxfs /dev/vx/rdsk/testdg/vol01
mount -F vxfs /dev/vx/dsk/testdg/vol01 /slavefs

after snapback,  snapvol-1 is gone. The plex will go back to vol01 for next split
vxassist -g testdg snapshot vol01 snapvol01
fsck -F vxfs /dev/vx/rdsk/testdg/snapvol01
mount -F vxfs /dev/vx/dsk/testdg/snapvol01 /snapfs/

7. remove snap
umount /snapfs
vxedit -g testdg -r -f rm snapvol01

chroot for ftp/telnet/rsh/ssh on solaris 10

By properly setting up a new root environment, you can fool the user and make them feel they are logging into a real root environment, rather than a jailed root. the funny part is to make a perfect root environment, with everything setting up, not exposure of the real /etc/passwd and /etc/shadow etc… The way to make a root environment I can think of includes 1. copy all required files (binary files, libraries, configurations …) down to the new root. 2. loopback mount the relevant file systems down to the new root. (the drawback is it will expose the real /etc to the user) 3. set the root path to the root path of a non-global zone. (make sure zone is running if you are running a sparse local zone)

 

1. ftp

It is fairly simple to get chroot for ftp session, change home directory from /etc/passwd as below,
usertest:55502:0:comments:/export/home/usertest/newroot/./test:/bin/ksh

/export/home/usertest/newroot/./test  is the new home directory, of which, /export/home/usertest/newroot is the new root, and /test is relateive path to the new root.

ftp server will change root to “/export/home/usertest/newroot/” after ftp session established.

Pls make sure you copy some essential required file down to the new root, such as /bin /sbin /etc,  etc.
the easy way is to loopback mount those file system as below

cd /export/home/usertest/newroot
mkdir bin sbin etc usr lib
mount -F lofs /bin  /export/home/usertest/newroot/bin
mount -F lofs /sbin  /export/home/usertest/newroot/sbin
mount -F lofs /usr  /export/home/usertest/newroot/usr
mount -F lofs /etc  /export/home/usertest/newroot/etc
mount -F lofs /lib  /export/home/usertest/newroot/lib

after user logs in via ftp, the session will be limited under the new root (/export/home/usertest/newroot)

pls note:  for ftp only, even the new root is total empty, the ftp session is still able to establish. But list functionality from ftp gui client doesn’t work, command line works well.
           authentication is based /etc/passwd and /etc/shadow under real root.

2.  telnet  (this example is based on solaris 10, change the procedure accordingly if you are running sol9 or sol
    it doesn’t matter if you do not change home directory in /etc/passwd

2.1 prepare the new root environment (/export/home/usertest/newroot)
    copy all required files/directories down to the new root and test it.
    eg:  /usr/sbin/chroot  /export/home/usertest/newroot  /bin/ksh

2.2
    inetadm -l telnet  > telnet.bak
    inetadm -m telnet exec=”/usr/sbin/a.ksh”
    svcadm restart telnet
   
    a.ksh
    #!/bin/ksh
    /usr/sbin/chroot /export/home/usertest/newroot/ /usr/sbin/in.telnetd
   
pls note, as chroot is explicitly run by a.ksh, the authentication will be based on /etc/passwd and /etc/shadow under new root, namely
/export/home/usertest/newroot/etc/passwd and /export/home/usertest/newroot/etc/shadow

3.  rsh
    it doesn’t matter if you do not change home directory in /etc/passwd
    rsh is using SMF svc:/network/login:rlogin and associated daemon is /usr/sbin/in.rlogind
   
3.1 prepare the new root
3.2 inetadm -m shell/rlogin exec=”/usr/sbin/b.ksh”

    b.ksh
    #!/bin/ksh
    /usr/sbin/chroot /export/home/usertest/newroot/ /usr/sbin/in.rlogind
   
pls note, authentication is based on /export/home/usertest/newroot/etc/passwd and /export/home/usertest/newroot/etc/shadow.

4.  openssh   

    authentication is based /etc/passwd and /etc/shadow under real root.
4.1  it is a must to change the home directory as above for ftp.
4.2  prepare the new root and test it.
4.3  download the below packages from sunfreeware.com and install them
     
     gcc
     libiconv
     openssl

4.4  download openssh source code and save into a temporary directory
     modify session.c as stated below

+ #define CHROOT                     /* to be added (of course without + mark at the beginning) */

  /* func */

  Session *session_new(void);
***************
*** 1159,1164 ****
— 1161,1171 —-
  void
  do_setusercontext(struct passwd *pw)
  {
+ #ifdef CHROOT                      /* to be added (of course without + mark at the beginning ) */
+       char *user_dir;              /* to be added (of course without + mark at the beginning) */
+       char *new_root;              /* to be added (of course without + mark at the beginning) */
+ #endif /* CHROOT */                /* to be added (of course without + mark at the beginning) */
+
        char tty=”;

  #ifdef HAVE_CYGWIN

                if (setlogin(pw->pw_name) < 0)
                        error(“setlogin failed: %s”, strerror(errno));

+               #ifdef CHROOT                  /* to be added (of course without + mark at the beginning) */
+                       user_dir = xstrdup(pw->pw_dir);  /* to be added (of course without + mark at the beginning) */
+                       new_root = user_dir + 1;  /* to be added (of course without + mark at the beginning) */
+
+                       while((new_root = strchr(new_root, ‘.’)) != NULL) {   /* to be added (of course without + mark at the beginning) */
+                               new_root–;                                   /* to be added (of course without + mark at the beginning) */
+                               if(strncmp(new_root, “/./”, 3) == 0) {        /* to be added (of course without + mark at the beginning) */
+                                       *new_root = ”;                     /* to be added (of course without + mark at the beginning) */
+                                       new_root += 2;                        /* to be added (of course without + mark at the beginning) */
+
+                                       if(chroot(user_dir) != 0)             /* to be added (of course without + mark at the beginning) */
+                                               fatal(“Couldn’t chroot to user directory %s”, user_dir);   /* to be added (of course without + mark at the beginning) */
+
+                                       pw->pw_dir = new_root;                 /* to be added (of course without + mark at the beginning) */
+                                       break;                                 /* to be added (of course without + mark at the beginning) */
+                               }                                              /* to be added (of course without + mark at the beginning) */
+                               new_root += 2;                                 /* to be added (of course without + mark at the beginning) */
+                       }                                                      /* to be added (of course without + mark at the beginning) */
+               #endif /* CHROOT */                                            /* to be added (of course without + mark at the beginning) */
+
                if (setgid(pw->pw_gid) < 0) {
                        perror(“setgid”);
                        exit(1);
     

4.5
./configure –with-random=/dev/random \
  –prefix=/usr/local –libexecdir=/usr/libexec/openssh \
  –sysconfdir=/usr/local/etc –mandir=/usr/share/man

4.6
make

4.7
mkdir /var/empty
chown root:sys /var/empty
chmod 755 /var/empty
groupadd sshd
useradd -g sshd -c ’sshd privsep’ -d /var/empty -s /bin/false sshd

4.8
make install

4.9
/usr/local/bin/ssh-keygen -t rsa1 -f /usr/local/etc/ssh_host_key -N
/usr/local/bin/ssh-keygen -t dsa -f /usr/local/etc/ssh_host_dsa_key -N
/usr/local/bin/ssh-keygen -t rsa -f /usr/local/etc/ssh_host_rsa_key -N

4.10
create file /etc/init.d/sshd

#!/bin/sh
pid=`/usr/bin/ps -e | /usr/bin/grep sshd |   /usr/bin/sed -e ’s/^  *//’ -e ’s/ .*//’`
case $1 in
’start’)
  /usr/local/sbin/sshd;;
’stop’)
  if [ "${pid}" != "" ]; then
  /usr/bin/kill ${pid}; fi;;
*)
  echo “usage: /etc/init.d/sshd (start|stop)”;;
esac

4.11
ln -s /etc/init.d/sshd /etc/rc2.d/S100sshd_chroot
ln -s /etc/init.d/sshd /etc/rc0.d/K100sshd_chroot

4.12
/etc/rc2.d/S100sshd_chroot start

problem with disk capacity

Yesterday, I was tasked to configure some new LUNs that allocated to a M500o server.  All luns worked very well except one.

syminq command showed the disk is with 18G space, luxadmin showed the same result as well.  But format command only recognize it as 9G.  

 

bash-3.00# syminq | grep -i 195E

/dev/rdsk/c3t60060480000290100478533031393545d0s2 M(2) EMC SYMMETRIX 5771 780195E008 17677440 (here 18G)

bash-3.00# luxadm display /dev/rdsk/c3t60060480000290100478533031393543d0s2

DEVICE PROPERTIES for disk: /dev/rdsk/c3t60060480000290100478533031393543d0s2

Vendor: EMC

Product ID: SYMMETRIX

Revision: 5771

Serial Num: 10047895C008

Unformatted capacity: 17263.125 Mbytes <—————–18G

Read Cache: Enabled

Minimum prefetch: 0×0

Maximum prefetch: 0xffff

Device Type: Disk device

Path(s):

/dev/rdsk/c3t60060480000290100478533031393543d0s2

/devices/scsi_vhci/ssd@g60060480000290100478533031393543:c,raw

Controller /devices/pci@12,600000/SUNW,qlc@0/fp@0,0

Device Address 5006048452a53fb9,6a

Host controller port WWN 2100001b320f2023

Class primary

State ONLINE

Controller /devices/pci@12,600000/SUNW,qlc@0,1/fp@0,0

Device Address 5006048452a53fb6,6a

Host controller port WWN 2101001b322f2023

Class primary

State ONLINE

 

But, with format

55. c3t60060480000290100478533031393545d0 <EMC-SYMMETRIX-5771 cyl 18412 alt 2 hd 15 sec 128>

/scsi_vhci/ssd@g60060480000290100478533031393545

56. c3t60060480000290100478533031393543d0 <EMC-SYMMETRIX-5771 cyl 9205 alt 2 hd 15 sec 128> default4 <— only 9205 cylinders

/scsi_vhci/ssd@g60060480000290100478533031393543

57. c3t60060480000290100478533031343033d0 <EMC-SYMMETRIX-5771 cyl 36826 alt 2 hd 30 sec 128>

/scsi_vhci/ssd@g60060480000290100478533031343033

partition> p

Volume: default4

Current partition table (original):

Total disk cylinders available: 9205 + 2 (reserved cylinders)

Part Tag Flag Cylinders Size Blocks

0 unassigned wu 0 0 (0/0/0) 0

1 unassigned wu 0 0 (0/0/0) 0

2 backup wu 0 – 9204 8.43GB (9205/0/0) 17673600 <—– 9G

3 unassigned wu 0 0 (0/0/0) 0

4 unassigned wu 0 0 (0/0/0) 0

5 unassigned wu 0 0 (0/0/0) 0

6 unassigned wu 0 0 (0/0/0) 0

7 unassigned wu 0 0 (0/0/0) 0

partition>

 

This is becuase the disk got wrongly labelled before, we just need to manually destroy the label and re-label it.

dd if=/dev/zero of=/dev/rdsk/c3t60060480000290100478533031393543d0s2 bs=512k count=1

then go back to format and re-label the disk,  afterwards, format reported it as 18G emc disk.

还是关于修改主机名，也要照顾VCS的感受！

上周末，成功将一个cluster中的两个节点替换了。原有系统是一个四节点的cluster，节点app1和app2  是两台 v480, db1和db2 是两台F6800。用两台M4000替换app1和app2。两台M4000分别命名为app11, app21，建立一个2节点的cluster，测试成功后，rename app11, app21 to app1，app2。 修改/etc/llttab /etc/llthosts,  相应的修改了main.cf 后，新节点加入后，一直显示节点app11加入，真是气人啊。

后来仔细看看文档， 发现还有个文件需要修改 /etc/VRTSvcs/conf/sysname, 如果该文件存在，cluster中的节点名字就来自它，如果不存在，就来在”uname -n”, 修改了这个文件后，成功将节点app1,app2重新加入到cluster中。

修改主机名，要考虑vxvm的感受

一些时候，需要修改主机名，对于vcs，相应的配置比较明了，不再累叙。对于vxvm，可以

1.  collect a copy of vxexplorer

2. vxdg -h <new-hostname> deport dgname 

    这步是修改vxvm disk上的信息

3. vxdctl hostid <newhostname>

    这步是修改/etc/vx/volboot 文件，以便将来有新盘加入的时候，写入新的hostname信息。

4. Change the hostname and IP address of the system

5. reboot server

但是在vxvm4.1 以及 5.0上，未做如上修改，主机重启后，disk groups 也都全部被自动import进来了。所以，估计除了hostname，vxvm肯定还是在disk上写了其他的信息，主机重启后，能自动识别disk 就是应该属于本机，就自动import进来了。

利用live upgrade 可以split 或者merge 文件系统

live upgrade除了可以用来在线升级操作系统，还可以用来split或者merge文件系统。

假如原有系统为solaris 10, 而且有两个硬盘 c0t0d0 ＆c0t1d0。 系统运行在 c0t0d0上，而且系统只有 根 （/） 和 swap 文件系统，现在想把 /usr 单独分离成一个文件系统，可以采取以下步骤分离

1. mount -F lofs /usr /usr

2. lucreate -c “sol10_old” -m /:/dev/dsk/c0t1d0s0:ufs \

                                                    -m -:/dev/dsk/c0t1d0s1:swap\

                                                    -m /usr:/dev/dsk/c0t1d0s3:ufs -n “sol10_new”

3. luactivate sol10_new

经过以上步骤就成功在c0t1d0上，将 /usr 分离出来，形成单独的文件系统了。 这里的技巧就是一定要 将 /usr mount成文件系统，因此 原来的/usr 只是个目录，因此可以用loopback 文件系统来愚弄 live upgrade

 

反之，我们也可以用 live upgrade 来merge文件系统。假如sol10运行在 c0t1d0上，有/  /usr swap 三个文件系统， 通过一下步骤，可以将/usr 合并到 / 中。

1. lucreate -c “sol1o_old” -m /:/dev/dsk/c0t0d0s0:ufs \

                                                   -m -:/dev/dsk/c0t0d0s1:swap -n “sol10_new”

2. luactivate sol10_new

VCS/VXVM upgrade & downgrade

公司的一个两节点的VCS 5.0  为了测试即将到来的应用系统升级，因此拿这个系统开刀，进行测试。虽然这个套系统一直还未投入使用，但上面也有1T的生产数据，所以操作还是很小心的，而且用BCV做了备份。

原有系统是solaris 9 + vcs/vxvm 5.0, 整个操作经过了以下阶段

1. vcs/vxvm downgrade to 4.1

2. 操作系统升级到 sol10 11/06 & vcs/vxvm 4.1 re-install

3. 操作系统升级到 sol10 05/08

4. vcs/vxvm 4.1 升级到 5.0mp1

整个过程波澜不惊，没什么可圈可点，只是第一步vcs/vxvm 从5.0 downgrade 到4.1 值得提一下。

vcs 的downgrade很简单，比较straightforward.

倒是vxvm 的downgrade比较tricky，因此veritas disk group 是不支持从5.0到4.1的downgrade，我特意与symantec 开了个case，他们也确认了这点。没办法，只有自己想招了。

无论4.1 还是 5.0 ， dg肯定还是不变的分为private ＆public region，只要保持privgate &public region 的大小，可以重建dg，来恢复数据，本着这个想法，经过摸索，终于成功讲vxvm downgrade 到4.1

1. backup vx dg information under v5

    vxprint -Qqmhspv -g sftg_datadg > dg1.txt

    verify configuration

    vxprint -D – -ht < dg1.txt

1.5

   check disk info

   vxprint -qd -g <dg_name>

 

 

2. destroy dg under vxvm v5.

    vxdg destory sftg_datadg

 

3. downgrade vxvm software from v5 to v4.1

    

4. Verify configuration under v4.1

vxprint -D – -ht < dg1.txt

 

delete those features that are not supported by v4.1 from dg1.txt, until dg1.txt is accepted by vxprint command

proxy_rid=0.1029

readonly=off

cons_reattach=off

fmr_rec_needed=off

voltype=off

siteconsistent=off

allsites=off

export=

site=

sd_name=

uber_name=

tentmv_src=off

tentmv_tgt=off

tentmv_pnd=off

 

5. create dg under vxvm 4.1

   vxdg -T 110 init sftg_datadg disk01=EMC0_1

 

5. recreate volume with configuration above

    vxmake -g sftg_datadg -d dg1.txt

6. start dg/vol

    vxvol -g sftg_datadg init active sftg

    mount /dev/vx/dsk/sftg_datadg/sftg /export/sftg

写在这里，便于自己查找，也希望能帮到其他人。

在solaris10 zone环境下部署VCS5

几个月前，自己设计并实施了solaris10 zone环境下VCS5的应用。整个系统框架如下：

系统为solaris10 05/08，两个节点，每个节点分别建3个zone，其中2个运行应用软件，1个运行oracle 10 数据库，运行数据库的zone内，运行三个oracle实例。应用程序和每个数据库实例都要求可以在两个节点内进行切换。集群软件采用veritas storage foundation5 MP1。 整个安装配置过程比较顺利，这里把需要注意的地方罗列一下
1. 由于三个local zone均采用exclusive-ip，也就是每个zone都分配了2个网卡，两个网卡都dedicated给相应的local zone。VCS5自带的IP agent只支持shared-ip local zone, 因此自己写了ZoneVip agent，通过zonecfg进行动态的配置ip。因为采用了exclusive-ip, local zone内有独立的ip stack，因此IPMP都是部署在local zone内，而且为了节省ip，采用link-based IPMP.

2. 由于VCS daemon必须运行在global zone内，而应用程序运行在local zone内，为了确保agent能够监测运行在local zone内的resource，而且在必要的时候进行相应的动作，inter-zone 的通信必要要建立起来，方法是在global zone内，建立好相应的vcs 用户，付好权限，在local zone内，首先通过环境变量VCS_HOST 指向global zone，然后执行 halogin username password, 进行登录操作，如果成功，会在local zone的跟下生成 .vcspwd 文件。

3. 一定要先建好local zone，挂接好 /opt 文件系统，然后在global zone内进行VCS/VXVM 安装，这样 veritas的相应package就都自动安装到了non-global zone 了。