Re: [SSI-users] iptables and CVIP?
Brought to you by:
brucewalker,
rogertsang
Menu
▾
▴
Re: [SSI-users] iptables and CVIP?
|
From: John B. <joh...@hp...> - 2005-02-05 01:25:48
|
Your cvip.conf is set to do NAT. Since both your nodes have access to the 10.1.242 network, you don't want to use NAT, you want to use DR. For DR, you do not need a gateway address, so delete that. That being said, I don't quite understand why your cluster is in the state is in. (The 10.1.242.246/32 is correct for an IP alias by the way.) There should be an IP alias on eth1 on the second node and there is not. Did you start ha-lvs by hand by any chance? Anyway, make the changes to your cvip.conf and reboot your cluster. ha-lvs should start during the boot and when node 2 joins. If you are trying to use iptables with cvip, I don't know how they will interact. So leave that off until cvip is working. John Byrne Jiann-Ming Su wrote: > How do I start iptables on the second node (it has an external interface)? > > Also, I'm trying to configure CVIP. I've read through the > README.CVIP, but I can't seem to get the cluster to respond to the > cvip. Here's what I have so far: > > # onall ip addr > 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > inet 127.0.0.1/8 brd 127.255.255.255 scope host lo > 2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 > link/ether 00:0f:1f:6c:ba:bb brd ff:ff:ff:ff:ff:ff > inet 192.168.0.1/24 brd 192.168.0.255 scope global eth0 > inet 10.1.242.2/32 scope global eth0 > 3: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 > link/ether 00:0f:1f:6c:ba:bc brd ff:ff:ff:ff:ff:ff > inet 10.1.242.25/24 brd 10.1.242.255 scope global eth1 > inet 10.1.242.246/32 scope global eth1 > (node 2) > 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > inet 127.0.0.1/8 brd 127.255.255.255 scope host lo > 2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 > link/ether 00:0f:1f:6c:ba:c5 brd ff:ff:ff:ff:ff:ff > inet 192.168.0.2/24 brd 192.168.0.255 scope global eth0 > 3: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 > link/ether 00:0f:1f:6c:ba:c6 brd ff:ff:ff:ff:ff:ff > inet 10.1.242.20/24 brd 10.1.242.255 scope global eth1 > > # cat /proc/cluster/lvs > CVIP Address Node number > 10.1.242.246 1 > # cat /proc/cluster/ip_vs_portweight > Port Range Weight > 1 - 80 1 > > # ipvsadm -Ln > IP Virtual Server version 1.0.10 (size=65536) > Prot LocalAddress:Port Scheduler Flags > -> RemoteAddress:Port Forward Weight ActiveConn InActConn > TCP 10.1.242.246:22 wlc > -> 192.168.0.2:22 Local 1 0 0 > > But, I can't connect to the cvip: > > $ telnet 10.1.242.246 22 > Trying 10.1.242.246... > > It doesn't respond to pings or traceroutes. But, I can access the > nodes themselves. > > $ telnet 10.1.242.20 22 > Trying 10.1.242.20... > Connected to node2.mydomain.bogus (10.1.242.20). > Escape character is '^]'. > SSH-1.99-OpenSSH_3.6.1p2 > > I thought it may have been an iptables problems, but I tried with > iptables turned off with the same results. What am I missing here? > The only thing I can tell is that cvip seems to think it's a /32 > instead of a /24 network. I've included my cvip.conf file as an > attachment. Thanks for any insight. > |