[SOLVED] keepalived can't ping virutal IP

bluethundr · 03-20-2014, 07:18 PM

Hey all,

I could really use some help with this one if anyone out there has this kind of experience.

I have keepalived installed on Red Hat Enterprise 5.7.

I have a very simple keepalived config that I'm attempting to use with HA/Proxy:

Code:

vrrp_instance VI_1 {
    state MASTER
    interface bond0
    virtual_router_id 50
    nopreempt
    priority 100
    advert_int 1
    virtual_ipaddress {
        192.168.200.11/24 dev eth2
    }
}

The service is able to run without any problem:

Code:

[root@uszmpdblp008la keepalived]# service keepalived start
Starting keepalived:                                       [  OK  ]

However I cannot ping the new virtual IP I am trying to create:

The ping attempt just hangs:

Code:

[root@uszmpdblp008la keepalived]# ping 192.168.200.11
PING 192.168.200.11 (192.168.200.11) 56(84) bytes of data.

ifconfig output does not show that the IP I am trying to create was ever created:

Code:

[root@uszmpdblp008la keepalived]# ifconfig
bond0     Link encap:Ethernet  HWaddr 2C:76:8A:56:F1:48
          inet addr:216.178.107.236  Bcast:216.178.107.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MASTER MULTICAST  MTU:1500  Metric:1
          RX packets:490894739 errors:48920 dropped:0 overruns:0 frame:2010651
          TX packets:63121734 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:40194144284 (37.4 GiB)  TX bytes:5804868240 (5.4 GiB)

eth0      Link encap:Ethernet  HWaddr 2C:76:8A:56:F1:48
          UP BROADCAST RUNNING SLAVE MULTICAST  MTU:1500  Metric:1
          RX packets:275003916 errors:24460 dropped:0 overruns:0 frame:1005326
          TX packets:63121734 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:22705347349 (21.1 GiB)  TX bytes:5804868240 (5.4 GiB)
          Interrupt:99 Memory:f6bf0000-f6c00000

eth2      Link encap:Ethernet  HWaddr 2C:76:8A:56:F1:48
          UP BROADCAST RUNNING SLAVE MULTICAST  MTU:1500  Metric:1
          RX packets:215890823 errors:24460 dropped:0 overruns:0 frame:1005325
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:17488796935 (16.2 GiB)  TX bytes:0 (0.0 b)
          Interrupt:99 Memory:f6b90000-f6ba0000

lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:5894514 errors:0 dropped:0 overruns:0 frame:0
          TX packets:5894514 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:1179798565 (1.0 GiB)  TX bytes:1179798565 (1.0 GiB)

So I am wondering if I am having a problem with keepalived because of the bonded interface. Because I generally haven't had any problem getting it to work in the past using non-bonded interfaces.

Any help or advice I can get on how to get this working correctly would be greatly appreciated.

bluethundr · 03-23-2014, 10:03 PM

Guys,

I'm not sure if anyone's ever checked out this thread. But I was actually able to figure this one out on my own! Basically, I remembered that our network in the datacenter was comprised of virtual machines (using VMWare virtualization) I wondered aloud if our network might bear some similarity to the Amazon EC2 network, which disallows broadcast/multicast transmissions? Some googling revealed that Willie Tarreau had come out with a patched version of keepalived that speaks unicast (“vrrp_unicast_bind” &”vrrp_unicast_peer“).

You can find that patched version here (even in RPM form):

http://1wt.eu/keepalived/

So this version was apparently generated in order to solve the problem of getting a heartbeat (using keepalived) solution to work in EC2. But that solution also worked on my company's network of virtual machines using VMWare. It's a nice solution that I hope can help someone else out here who might be having the same problem.

Cheers!