VIP loopback rc.conf

foudebassan · 2010-05-11 07:58:38

Hello everyone,

I'm trying to set up a load-balanced website using keepalived.
I have to set up a virtual ip on the loopback interface on each web node.

I'm trying this syntax in the rc.conf file :

eth1="eth1 192.168.41.253 netmask 255.255.254.0 broadcast 192.168.41.255"
lo_1="lo:1 192.168.41.250 netmask 255.255.255.255 broadcast 192.168.41.250"
INTERFACES=(eth1 lo_1)

On the machine startup, i get this message :
SIOCSIFFLAGS Cannot assign requested address

The interfaces seems ok, but i would like to have a clean startup.
When i restart the network, i always have a failure on network shutdown, but the startup is ok.

eth1 Link encap:Ethernet HWaddr 08:00:27:14:E9:07
inet addr:192.168.41.252 Bcast:192.168.41.255 Mask:255.255.254.0
inet6 addr: fe80::a00:27ff:fe14:e907/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:17 errors:0 dropped:0 overruns:0 frame:0
TX packets:875 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1020 (1020.0 b) TX bytes:47298 (46.1 Kb)
Interrupt:9 Base address:0xd240

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:61 errors:0 dropped:0 overruns:0 frame:0
TX packets:61 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:9806 (9.5 Kb) TX bytes:9806 (9.5 Kb)

lo:1 Link encap:Local Loopback
inet addr:192.168.41.250 Mask:255.255.255.255
UP LOOPBACK RUNNING MTU:16436 Metric:1

[root@hanzel ~]# /etc/rc.d/network restart
:: Stopping Network [FAIL]
:: Starting Network [DONE]
[root@hanzel ~]#ifconfig
eth1 Link encap:Ethernet HWaddr 08:00:27:14:E9:07
inet addr:192.168.41.252 Bcast:192.168.41.255 Mask:255.255.254.0
inet6 addr: fe80::a00:27ff:fe14:e907/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:26 errors:0 dropped:0 overruns:0 frame:0
TX packets:943 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1560 (1.5 Kb) TX bytes:50994 (49.7 Kb)
Interrupt:9 Base address:0xd240

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:192 errors:0 dropped:0 overruns:0 frame:0
TX packets:192 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:22826 (22.2 Kb) TX bytes:22826 (22.2 Kb)

lo:1 Link encap:Local Loopback
inet addr:192.168.41.250 Mask:255.255.255.255
UP LOOPBACK RUNNING MTU:16436 Metric:1

Is there another syntax in the rc.conf file in order to remove theses failure messages.

Last edited by foudebassan (2010-05-11 08:56:03)

fukawi2 · 2010-05-11 08:34:34

foudebassan wrote:

eth1="eth1 192.168.41.253 netmask 255.255.254.0 broadcast 192.168.41.255"
lo_1="lo:1 192.168.41.250 netmask 255.255.255.255 broadcast 192.168.41.250"
INTERFACES=(eth1 l0_1)

1) I'm guessing you want lo_1 rather than l0_1
2) You shouldn't need to assign an RFC1918 address to lo... Ever AFAIK. You certainly don't need to for keepalived

I run several H/A box pairs at my work with keepalive and we don't have anything other than 127.0.0.1 on lo

Last edited by fukawi2 (2010-05-11 08:35:20)

Sin.citadel · 2010-05-11 08:44:14

i checked this config myself and i am also getting the fail message, but since startup is fine, it isnt a concerns, although if you dont want to see this message, you can add the virtual interface line to rc.local and run it from there.

foudebassan · 2010-05-11 09:03:13

Thanks for testing Sin.citadel.

fukawi2,
I mistyped "l0_1" in my first post, i corrected it.
I think you use keepalived doing NAT, that's why you dont use the loopback interface.

In my network configuration, my LVS and my web nodes are on the same subnet. I'm using direct routing in my keepalived configuration.
Using NAT is useless for me, and requires more CPU ressources.
When using DR, the VIP is shared by each node on the loopback interface, with some tuning in the sysctl.conf file to correct arp issues, it runs fine.

I was just looking for a way to have a clean boot. Using the rc.local file could be a solution

Last edited by foudebassan (2010-05-11 09:07:56)

fukawi2 · 2010-05-11 22:57:36

I'm not sure what you mean by using it for NAT... We use it for H/A configurations on firewall perimeter devices, so the boxes do perform NAT, but they also provide proxy, SMTP relay etc via the keepalived virtual address.

Arch Linux

#1 2010-05-11 07:58:38

VIP loopback rc.conf

#2 2010-05-11 08:34:34

Re: VIP loopback rc.conf

#3 2010-05-11 08:44:14

Re: VIP loopback rc.conf

#4 2010-05-11 09:03:13

Re: VIP loopback rc.conf

#5 2010-05-11 22:57:36

Re: VIP loopback rc.conf

Board footer