Comment 7 for bug 717166

Revision history for this message
Daniel Nurmi (nurmi) wrote :

Carlos,

This is the message that was being thrown when the original problem was cropping up; if you're still seeing this message with the patch installed, then it (the patch) is not working properly. The server should run even though there is no address in the config on the 10.55.55.0/24 subnet, since there should be another IP from the 172.19 subnet assigned to one of the eth0* interfaces. Can you please verify that the patch is applied, and perhaps show the output of 'ip addr show' when this error is occurring? I've tested again on my local setup with the patched dhcpd and it is launching successfully. It should look something like this, where

169.254.169.254/32 is the metadata service redirect IP
192.168.6.148/18 is the physical interface IP (analogous to your 10.55.55.0/24 subnet)
192.168.16.40/32 is one of the VNET_PUBLICIP addresses specified in eucalyptus.conf
172.19.1.1/27 is drawn from the VNET_SUBNET/VNET_NETMASK network specified in eucalyptus.conf

2: eth4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether d0:aa:aa:bb:cc:dd brd ff:ff:ff:ff:ff:ff
    inet 169.254.169.254/32 scope link eth4
    inet 192.168.6.148/18 brd 192.168.63.255 scope global eth4
    inet 192.168.16.40/32 scope global eth4
    inet 172.19.1.1/27 brd 172.19.1.31 scope global eth4:priv
    inet6 fe80::d2aa:aaff:febb:ccdd/64 scope link
       valid_lft forever preferred_lft forever

root@eucahost-4-243:~# ps ax | grep dhcpd
 4519 ? Ss 0:00 /usr/sbin/dhcpd -cf //var/run/eucalyptus/net/euca-dhcp.conf -lf //var/run/eucalyptus/net/euca-dhcp.leases -pf //var/run/eucalyptus/net/euca-dhcp.pid -tf //var/run/eucalyptus/net/euca-dhcp.trace eth4