1、配置用户接入VLAN100，连接用户电脑/STB。
vlan 100
  name STB_PC
  multicast-packet drop-unknown

2、IGMP相关参数检查。缺省情况下，IGMP都是起来的，不用额外配置。
igmp enable
igmp version-mode v1 accept
igmp version-mode v2 accept
igmp version-mode v3 accept
igmp span-vlan enable
igmp snooping-aging-time 300
igmp bandwidth-control disable
igmp host-tracking disable
igmp robustness 2
igmp query-interval 125
igmp query-max-resp 100
igmp last-query-interval 10
igmp last-query-count 2
igmp v1router-timeout 400
igmp unsolicited-report-interval 10
igmp non-match-group drop
igmp mcm disable

3、配置MVLAN。MVLAN必须配置，且须与上联路由器或三层交换机所配置的VLAN号相同，亦即VLAN 100。IGMP模式可以是proxy，也可以是snooping，推荐proxy模式，据中兴说工作稳定些。将组播组239.192.0.1加入到MVLAN，也可以同时加多个组播组，并定义所用的带宽，这里限定为10M。
igmp mvlan 100
igmp mvlan 100 enable
igmp mvlan 100 work-mode proxy
igmp mvlan 100 host-ip 0.0.0.0
igmp mvlan 100 max-group 16
igmp mvlan 100 group-filter enable
igmp mvlan 100 group 239.192.0.1
igmp mvlan 100 group 239.192.0.1 bandwidth 10240
//也可以用下面命令指定该组播组的源IP地址
igmp mvlan 100 group 239.192.0.1  source-address 192.168.0.1

4、OLT上联千兆端口为Trunk口，VLAN100透传到上级接入交换机，接入交换机与OLT互联端口也为Trunk，VLAN100的网关在上级三层交换机上。
interface gei_0/2/1
  hybrid-attribute fiber
  negotiation auto
  flowcontrol disable
  port-protect disable
  switchport mode trunk
  switchport vlan 1,100 tag

5、OLT PON口配置，ONU须先注册好。
interface epon-olt_0/11/1
  no shutdown
  p2p mode group
  onu 1 type ZTE-D402 mac 001e.738f.884c ip-cfg static

6、ONU配置，ONU型号为D402，其中第1口连接STB，并接收组播数据，ONU缺省是IGMP Snooping模式。
interface epon-onu_0/11/1:1
  switchport mode trunk vport 1
  switchport vlan 1,100  tag vport 1
  authentication enable
  description $1$测试1$$
  spliter 1 line 1
  bandwidth upstream fixed 0 fixedpkt 64 assured 10000 maximum 10000 maxburst 45000 priority 0
  bandwidth downstream maximum 10000 maxburst 205000
  encrypt direction downstream  enable
  autocfg disable

pon-onu-mng epon-onu_0/11/1:1
  no Forbid-auto-dispatch
  Forbid-auto-dispatch voip
  vlan port eth_0/1 mode tag vlan 100 priority 0
  multicast vlan port eth_0/1 add vlanlist 100   //端口须加入组播VLAN
  multicast vlan tag-strip port eth_0/1 enable   //去掉VLAN tag
  multicast group-max-number eth_0/1 16          //可配置最大组播组数量

7、设置组播的源端口和接收端口，源端口为上联千兆端口，收到端口为需要接收组播的ONU PON口。
igmp mvlan 100 source-port gei_0/2/1
igmp mvlan 100 receive-port epon-onu_0/11/1:1 vport 1

8、当ONU下连不是电脑或STB，而是连接三层交换机，并且该三层交换机需与上级交换机建立PIM邻居关系时，由于ONU缺省并不转发224.0.0.x组播组，从而导致无法建立PIM邻居关系。这时需要将ONU配置为IGMP Proxy模式。实际上OLT对该模式支持不好，当ONU加电在线时，是无法配置的，只有不在线时可以配置，另外通过命令行也无法配置，在命令行下，multicast switch命令后面必须跟control-multicast或igmpsnooping，如下所示：
ZXA10(epon-onu-mng)#multicast switch ?
  control-multicast  Control Multicast
  igmpsnooping       IGMP snooping

当成功设置ONU工作模式后，可以后ONU的配置信息如下,出现了multicast switch命令。
pon-onu-mng epon-onu_0/11/1:1
  multicast switch
  no Forbid-auto-dispatch
  Forbid-auto-dispatch voip
  vlan port eth_0/1 mode tag vlan 100 priority 0
注：可能OLT软件版本对组播支持不够好，才会出现这种怪异的配置现象。

9、在OLT上可以采用以下命令进行测试。
show igmp              //显示IGMP信息
show igmp mvlan        //显示MVLAN基本状态信息
show igmp mvlan  100   //显示MVLAN中Source/Receive端口等信息
show igmp mvlan  100 group 239.192.0.1  //显示在接收该组播的ONU
show remote onu multicast-info epon-onu_0/11/1:1    //显示ONU组播配置信息

 number of unicast mac addresses:   5K
 number of igmp groups:             1K
 number of qos aces:                512
 number of security aces:           512
 number of unicast routes:          8K
 number of multicast routes:        1K

2、MVRF配置
ip vrf test
 rd 100:1
 route-target export 100:1
 route-target import 100:1

ip multicast-routing
ip multicast-routing vrf test
ip pim vrf test rp-address 192.168.0.254

interface Vlan10
 description "Uplink vlan"
 ip vrf forwarding test
 ip address 192.168.10.1 255.255.255.0
 ip pim sparse-mode

interface Vlan100
 description "Users Vlan"
 ip vrf forwarding test
 ip address 192.168.100.254 255.255.255.0
 ip pim sparse-mode

3、问题处理
   (1) 起初配置时，以为象Catalyst3750G一样，只需在VRF内启用multicast-routing，配置完成后，在 vlan100上配置ip igmp join-group 239.192.0.1命令测试组播，发现无法建立source Tree，相关命令查看信息如下（注：下述全部略去(*, 224.0.1.40)条目相关信息）：
router#show ip mroute vrf test
IP Multicast Routing Table
Flags: D - Dense, S - Sparse, B - Bidir Group, s - SSM Group, C - Connected,
       L - Local, P - Pruned, R - RP-bit set, F - Register flag,
       T - SPT-bit set, J - Join SPT, M - MSDP created entry,
       X - Proxy Join Timer Running, A - Candidate for MSDP Advertisement,
       U - URD, I - Received Source Specific Host Report,
       Z - Multicast Tunnel, z - MDT-data group sender,
       Y - Joined MDT-data group, y - Sending to MDT-data group
       V - RD & Vector, v - Vector
Outgoing interface flags: H - Hardware switched, A - Assert winner
 Timers: Uptime/Expires
 Interface state: Interface, Next-Hop or VCD, State/Mode

(*, 239.192.0.1), 00:02:55/00:02:36, RP 192.168.0.254, flags: SJCL
  Incoming interface: Vlan10, RPF nbr 192.168.10.254
  Outgoing interface list:
    Vlan100, Forward/Sparse, 00:02:55/00:02:36
从上可以看出，mroute中无(S,G)条目，亦即未建立Source Tree。

router#show ip mroute vrf test count
IP Multicast Statistics
2 routes using 1140 bytes of memory
2 groups, 0.00 average sources per group
Forwarding Counts: Pkt Count/Pkts(neg(-) = Drops) per second/Avg Pkt Size/Kilobits per second
Other counts: Total/RPF failed/Other drops(OIF-null, rate-limit etc)

Group: 239.192.0.1, Source count: 0, Packets forwarded: 0, Packets received: 0

router#show ip mroute vrf test active                  
Active IP Multicast Sources - sending >= 4 kbps
 a negative (-) Rate counts pps being fast-dropped

router#show ip pim vrf test interface vlan 100 count

State: * - Fast Switched, D - Distributed Fast Switched
       H - Hardware Switching Enabled
Address          Interface                FS  Mpackets In/Out
192.168.100.254   Vlan100                  * H 0/0
从上面几条命令显示结果可以看出，交换机并未接收到组播数据，即使mroute中有(*, 239.192.0.1)条目，并加入了RP-tree，但由于上级PE通过Source Tree发布组播数据，所以未收到。

    (2) 全局配置ip multicast-routing命令后，mroute中就有了（S,G）条目，也能正常收到组播数据了。相关命令显示信息如下：
router#show ip mroute vrf test       
IP Multicast Routing Table
Flags: D - Dense, S - Sparse, B - Bidir Group, s - SSM Group, C - Connected,
       L - Local, P - Pruned, R - RP-bit set, F - Register flag,
       T - SPT-bit set, J - Join SPT, M - MSDP created entry,
       X - Proxy Join Timer Running, A - Candidate for MSDP Advertisement,
       U - URD, I - Received Source Specific Host Report,
       Z - Multicast Tunnel, z - MDT-data group sender,
       Y - Joined MDT-data group, y - Sending to MDT-data group
       V - RD & Vector, v - Vector
Outgoing interface flags: H - Hardware switched, A - Assert winner
 Timers: Uptime/Expires
 Interface state: Interface, Next-Hop or VCD, State/Mode

(*, 239.192.0.1), 00:15:17/stopped, RP 192.168.0.254, flags: SJCL
  Incoming interface: Vlan10, RPF nbr 192.168.10.254
  Outgoing interface list:
    Vlan100, Forward/Sparse, 00:15:17/00:02:19

(192.168.1.1, 239.192.0.1), 00:00:07/00:02:59, flags: LJT
  Incoming interface: Vlan10, RPF nbr 192.168.10.254
  Outgoing interface list:
    Vlan100, Forward/Sparse, 00:00:07/00:02:52

router#show ip mroute vrf test active
Active IP Multicast Sources - sending >= 4 kbps
 a negative (-) Rate counts pps being fast-dropped

Group: 239.192.0.1, (?)
   Source: 192.168.1.1 (?)
     Rate: 295 pps/3131 kbps(1sec), 2984 kbps(last 40 secs), 3269 kbps(life avg)

router#show ip mroute vrf test count 
IP Multicast Statistics
3 routes using 1604 bytes of memory
2 groups, 0.50 average sources per group
Forwarding Counts: Pkt Count/Pkts(neg(-) = Drops) per second/Avg Pkt Size/Kilobits per second
Other counts: Total/RPF failed/Other drops(OIF-null, rate-limit etc)

Group: 239.192.0.1, Source count: 1, Packets forwarded: 18648, Packets received: 18648
  RP-tree: Forwarding: 0/0/0/0, Other: 0/0/0
  Source: 192.168.1.1/32, Forwarding: 18648/288/1325/3055, Other: 18648/0/0

router#show ip pim vrf test interface vlan 100 count

State: * - Fast Switched, D - Distributed Fast Switched
       H - Hardware Switching Enabled
Address          Interface                FS  Mpackets In/Out
192.168.100.254   Vlan100                  * H 0/31119

2、安装前检查操作系统是否满足安装要求
-bash-3.00$ cd install-scripts/
-bash-3.00$ ls
check_prerequisites.sh  dbmig_21_250.sh   dbrename_25_300.sh  rpt-tmpl-scas-modifier.exe
check_syb.sh            dbmig_250_255.sh  install-cm.sh       set_shmmax.sh
-bash-3.00$ pwd
/usr/tmp/cm_install_temp/install-scripts
-bash-3.00$ ./check_prerequisites.sh
Check: Minimal CPU Speed: 800
         [Success] CPU Speed is sufficient: 3192MHz

Check: Minimal OS Version: RHEL3/4
         [Success] Found RHEL4, Checking minimal update level:
         [Success] Update level: 0

Check: Additional RPMS
         [Failed] Installed version is "compat-libstdc++-33-3.2.3-47.3"
         [Notice] If installed version is greater than expected, you may ignore this error

Check: Total memory
         [Success] Found 1034488 KB memory

Check: Max Shared memory
         [Success] Found 1073741824 bytes of shared memory

Check: Python in execution path
         [Success] Python found at /usr/bin/python

Check: 8096000K free on intended CM home dir:
         [Warning] No CM home directory given. Cannot verify there is enough free space to install CM

Check: 1024000K free on intended Sybase home dir:
         [Warning] No sybase home directory given. Cannot verify there is enough free space to install Sybase

Check: 15000000K free on intended Sybase Data dir:
         [Warning] No sybase Data directory given.
         [Notice] Only applicable if you intend on using Datadir installation method

Check: Locale set to en_US
         [Success] System Locale set to en_US
         [Success] Current Locale set to en_US

Check: Names for all NIC's
         [Success] IP Address 202.158.161.13 is resolvable through /etc/hosts
         [Success] IP Address 127.0.0.1 is resolvable through /etc/hosts

3、设置kernel.shmmax。通过修改/etc/sysctl.conf达到此目的。在该文件内添加以下一行
  kernel.shmmax = 1073741824
  这个更改在系统重新启动后生效。若不设置该参数，在安装Sybase数据库时会出现以下错误：
  [root@localhost sybase]# ./installsyb.sh --sybhome=/usr/sybhome --datadir=/usr/sybdata
  Starting Sybase install sequence
  testing dist file integrity...done
  Running platform-specific check file...found 33554432 bytes shared memory, need at least 512000000 bytes (put new value in /proc/sys/kernel/shmmax, and make permanent by setting 'kernel.shmmax' in /etc/sysctl.conf)
  Error - platform-specific check failed - exiting.

4、安装Sybase数据库。放在SYBASE目录下安装，不要在全局下安装，会出错，sybdata要提前建立sybhome不要建立。首先以root用户登录到系统，进入软件安装目录：
  [root@scecm sybase]# ls
  common.sh  config  installsyb.sh  Linux  syb_interfaces.sh
  [root@scecm sybase]# pwd
  /usr/tmp/cm_install_temp/sybase

  开始安装SYBASE
  [root@localhost sybase]# ./installsyb.sh --sybhome=/usr/sybhome --datadir=/usr/sybdata  Starting Sybase install sequence
  testing dist file integrity...done
  Running platform-specific check file...done
  Adding sybase user...done
  NOTE: if this installation aborts or fails, you may want
  to remove the sybase process/user before attempting again.  This
  can be done with the command:
                     pkill -9 -xU sybase; userdel -r sybase
     
  Extracting dist file to /usr/tmp/sybinstall29739...done
  Running Sybase installer - this will take a while...done
...
...
  在安装过程中提示设置保存SCE监控数据的数据库大小，即文件/usr/sybdata/apticotdata.dat的大小，这里设置为30G左右，保留近2月的数据
  Waiting for Sybase to come up... (0 of 60)
  Please enter SIZE in 2K blocks of file to be used for the "data[maximum is 41661909; minimum is 5242798]" device.
  NOTE - the actual size required will include a 1.05 overhead on the amount you specify.
  SIZE in 2K blocks: 15728394
  Do you want to input the values again? (yes/no): no
  Password correctly set.
...
...
  Skipping file configsyb.sh
  Skipping file sybase-etc-init-script
  /usr/tmp/cm_install_temp/sybase/config
  Done Sybase configuration sequence
  done
  cleaning up...done.
  至此Sybase数据库安装结束。

5、安装CM3.1.6
  [root@scecm install-scripts]# ls
  check_prerequisites.sh  dbmig_21_250.sh   dbrename_25_300.sh  rpt-tmpl-scas-modifier.exe
  check_syb.sh            dbmig_250_255.sh  install-cm.sh       set_shmmax.sh
  [root@scecm install-scripts]# pwd
  /usr/tmp/cm_install_temp/install-scripts
 
  检查Sybase数据库安装情况
  [root@scecm install-scripts]# ./check_syb.sh
  Sybase is properly installed and configured

  开始安装CM
  [root@localhost install-scripts]# ./install-cm.sh -d /usr/cm
  Adding/verifying user and group for SCMS-CM ...
  Populating scmscm home /usr/cm ...
  Populating scmscm home /usr/cm from  ../HomeCM.tgz ...
  ...
  ...
  Making symlinks in /etc/init.d and rc*.d  ...
  done
  Done installation.
  NOTE: The user name created for SCMS-CM is scmscm.  You may want to set a password for this account now.

6、设置账号scmscm密码
  [root@localhost install-scripts]# pas swd scmscm
  Changing password for user scmscm.
  New UNIX password:
  BAD PASSWORD: it does not contain enough DIFFERENT characters
  Retype new UNIX password:
  passwd: all authentication tokens updated successfully.

7、以scmscm账号登录设置系统时区
  -bash-3.00$ ls
  clumain.sh  cm  cm.jar  jselect-sce-tz.sh  p3rpc  update_vlinks.sh
  -bash-3.00$ pwd
  /usr/cm/cm/bin
  -bash-3.00$ ./jselect-sce-tz.sh --offset=480
                TIME_STAMP|              OFFSET_MIN|
  ------------------------+------------------------+
   2008-10-07 13:37:18.256|                     480|

8、起CM，查看工作状态，并可根据情况设置一些参数
  -bash-3.00$ ./p3rpc --set-user --username=scmscm --password=scmscm
  Command terminated successfully

  -bash-3.00$ ./cm start
  Starting CM:

  -bash-3.00$ cd setup/
  -bash-3.00$ ls
  alive.sh  common.sh  mbean.py  monitor  on-boot.py  RUN_0_cm  scmscm  STOP_1_cm  sybase  syb_interfaces.sh
  -bash-3.00$ ./alive.sh     //查看CM运行状态
  STATUS OK
  -bash-3.00$ ./on-boot.py    //查看当前的CM和Sybase startup状态
  sybase:  on
  cm:  on

  -bash-3.00$ cd scripts/
  -bash-3.00$ pwd
  /usr/cm/scripts
  -bash-3.00$ ls
  adapterconf.py  csvconf.sh  dbperiodic.py  droptable.sh   rdr-rate.py  sybback.sh
  common.sh       dbfree.sh   dbtables.sh    prunetable.sh  sceconf.sh   sybrestore.sh
  -bash-3.00$ ./dbfree.sh     //查看数据库使用情况
  Name                    % Data Free % Log Free
  ---------------------- ----------- ----------
  Database                        96      99
  -bash-3.00$ ./dbtables.sh   
  Executing query ...
                      name|               num_lines|                min_time|                max_time|
  ------------------------+------------------------+------------------------+------------------------+
                   RPT_SUR|                       0|                    null|                    null|
                   RPT_LUR|                  101270| 2008-10-03 10:49:11.486|  2008-10-05 18:45:33.69|
                  RPT_VLUR|                       0|                    null|                    null|
                   RPT_PUR|                   52030| 2008-10-03 10:49:11.503| 2008-10-05 18:45:38.676|
                    RPT_TR|                 8106260|  2008-10-03 10:49:10.84|  2008-10-05 18:47:10.82|
                 RPT_MALUR|                  274710|  2008-10-03 10:49:10.84| 2008-10-05 18:47:03.836|
                 RPT_MEDIA|                   10000| 2008-10-03 10:49:10.853|  2008-10-05 18:46:13.17|
          RPT_TOPS_PERIOD0|                   16335| 2008-10-03 11:00:00.003| 2008-10-07 13:00:00.183|
          RPT_TOPS_PERIOD1|                     660|  2008-10-04 00:00:00.02| 2008-10-07 00:00:00.163|

-bash-3.00$ ./sceconf.sh --op=list   //查看SCE
IP              Rate            Peak
------------    -----------     -----------
×.×.×.×          0.0             47.050785

Command terminated successfully

其它问题可以参考http://www.cnitblog.com/zty/archive/2008/10/03/49730.html