Recently i came acrossed a server , in which IPMP is configured entirely in a different way .
Normally to configure IPMP, we need atleast 2 Physical interface . Each interface needs to have dedicated IPaddress and one virtual ipaddress needs to be configured to perform failover between the 2 nics .
But the server i came across has following configuration in their hostname.e100g* files and the IP address 193.XX.XX.15 is configured only ong the e1000g0 . When i disconnect a nic card in e1000g0, the ipaddress is failling over to e1000g1. I have no idea how they configured , their is no "addif" command in the /etc/hostnames.e100g* files. No virtual ipaddress configured . I am puzzled how the nic failover works. I just want to know how IPMP configured in this server. Please advise me , if i need to attach any other information.
lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
e1000g0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
inet 193.xx.xx15 netmask ffffff00 broadcast 193.xx.xx.255
e1000g1: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 3
inet 0.0.0.0 netmask ff000000 broadcast 0.255.255.255
monappp068 group ipmp_grp up
group ipmp_grp up
You can configure IPMP in two ways:
- probe based: you configure one IP in each nic and one "virtual IP". The probes on the interfaces are made by the ipmp daemon pinging the default router.
- link based: you configure only one ip on one nic and define the ipmp group on all the nic members. When the ipmp daemon detects a link-lost signal on the physical media, produces a failover to the "backup" nic.