同网段多套sun cluster，系统管理的心跳地址出现冲突如何解决？

zhiwei.ma

一个网段内 两套sun cluster后 scstat看起来cluster完全正常，日志出现心跳ip地址冲突的提示。经查冲突ip的mac确实是另一套双机上的心跳地址的mac。如何解决？会不会影响cluster的运行？

hosts
#vi /etc/hosts

127.0.0.1       localhost
192.168.1.41   SC1     loghost
192.168.1.43   SC1-e1000g0-test
192.168.1.44   SC1-nxge0-test
192.168.1.42   SC2
192.168.1.45   SC2-e1000g0-test
192.168.1.46   SC2-nxge0-test

192.168.1.32 ora-lh
192.168.1.33 yilian-lh


主机ipmp:
#vi /etc/hostname.e1000g0
SC1 mtu 1450 netmask + broadcast + group sc_ipmp0 up addif SC1-e1000g0-test mtu 1450 netmask + broadcast + deprecated -failover up

#vi /etc/hostname.nxge0
SC1-nxge0-test mtu 1450 netmask + broadcast + group sc_ipmp0 deprecated -failover standby up

备机ipmp
#vi /etc/hostname.e1000g0
SC2 mtu 1450 netmask + broadcast + group sc_ipmp0 up addif SC2-e1000g0-test mtu 1450 netmask + broadcast + deprecated -failover up

#vi /etc/hostname.nxge0
SC2-nxge0-test mtu 1450 netmask + broadcast + group sc_ipmp0 deprecated -failover standby up

[ 本帖最后由 zhiwei.ma 于 2008-12-18 08:49 编辑 ]

welcome008

这个应该不影响吧？

那两个网卡不是直连的吗？

如果担心的话，就改地址算了，不过可能需要停业务了

我猜的

Snowict

修改心跳地址，比较好。

zhiwei.ma

#vi /etc/hostname.e1000g0
SC1 mtu 1450 netmask + broadcast + group sc_ipmp0 up \
addif SC1-e1000g0-test mtu 1450 netmask + broadcast + deprecated -failover up
和下面
#vi /etc/hostname.e1000g0
SC1 mtu 1450 netmask + broadcast + group sc_ipmp0 up addif SC1-e1000g0-test mtu 1450 netmask + broadcast + deprecated -failover up
有没有区别？与此有关吗？

wind521

网络上把那几个划到不同的VLAN里

zhiwei.ma

root@SC1 # ifconfig -a
lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index
1
        inet 127.0.0.1 netmask ff000000
e1000g0: flags=1001000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4,FIXEDMTU> mtu 1450
index 2
        inet 192.168.1.43 netmask ffffff00 broadcast 192.168.1.255
        groupname sc_ipmp0
        ether 0:14:4f:c4:99:6a
e1000g0:1: flags=1009040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,NOFAI
LOVER,FIXEDMTU> mtu 1450 index 2
        inet 192.168.1.45 netmask ffffff00 broadcast 192.168.1.255
e1000g0:2: flags=1040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4> mtu 150
0 index 2
        inet 192.168.1.32 netmask ffffff00 broadcast 192.168.1.255
e1000g0:3: flags=1040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4> mtu 150
0 index 2
        inet 192.168.1.33 netmask ffffff00 broadcast 192.168.1.255
e1000g1: flags=1008843<UP,BROADCAST,RUNNING,MULTICAST,PRIVATE,IPv4> mtu 1500 ind
ex 5
        inet 172.16.0.129 netmask ffffff80 broadcast 172.16.0.255
        ether 0:14:4f:c4:99:6b
nxge0: flags=1069040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,NOFAILOVE
R,STANDBY,INACTIVE,FIXEDMTU> mtu 1450 index 3
        inet 192.168.1.46 netmask ffffff00 broadcast 192.168.1.255
        groupname sc_ipmp0
        ether 0:14:4f:c1:48:8
nxge1: flags=1008843<UP,BROADCAST,RUNNING,MULTICAST,PRIVATE,IPv4> mtu 1500 index
4
        inet 172.16.1.1 netmask ffffff80 broadcast 172.16.1.127
        ether 0:14:4f:c1:48:9
clprivnet0: flags=1009843<UP,BROADCAST,RUNNING,MULTICAST,MULTI_BCAST,PRIVATE,IPv
4> mtu 1500 index 6
        inet 172.16.193.1 netmask ffffff00 broadcast 172.16.193.255
        ether 0:0:0:0:0:1
root@SC1 #

[ 本帖最后由 zhiwei.ma 于 2008-12-18 09:14 编辑 ]

zhiwei.ma

现在应用都正常切主备机也正常，可不可以先将资源切到一台机然后降另一台先停掉，等交换机配好valn再起来？有个担心这时候跑资源的主机能不能发现这台已经规划到valn的备机？
将心跳单独划到valn里会不会影响cluster？

[ 本帖最后由 zhiwei.ma 于 2008-12-18 12:27 编辑 ]

zhiwei.ma

Dec 18 04:37:07 SC1 ip: [ID 876157 kern.warning] WARNING: node 00:14:4f:1f:03:ab
is suing our IP address 172.016.000.129 on nxge0
Dec 18 04:37:07 SC1 ip: [ID 876157 kern.warning] WARNING: node 00:14:4f:1f:03:ab
is suing our IP address 172.016.000.129 on e1000g0
Dec 18 04:47:07 SC1 ip: [ID 876157 kern.warning] WARNING: node 00:14:4f:1f:03:ab
is suing our IP address 172.016.000.129 on nxge0
Dec 18 04:37:07 SC1 ip: [ID 876157 kern.warning] WARNING: node 00:14:4f:1f:03:ab
is suing our IP address 172.016.000.129 on e1000g0
......
.....
....


就这样报下去 每十分钟一次

[ 本帖最后由 zhiwei.ma 于 2008-12-18 09:42 编辑 ]

huan.yang

麻烦把报错信息贴出来

zhiwei.ma

原帖由 huan.yang 于 2008-12-18 09:33 发表
麻烦把报错信息贴出来

在你楼上