ISC DHCP on a Solaris 10 Container

Glenn Satchell Glenn.Satchell at uniq.com.au
Wed Mar 15 15:16:01 UTC 2006


>Date: Mon, 13 Mar 2006 16:48:04 +0100
>From: Pascal Gienger <Pascal.Gienger at uni-konstanz.de>
>To: dhcp-users at isc.org
>Subject: Re: ISC DHCP on a Solaris 10 Container
>
>Mike Diggins <diggins at McMaster.CA> wrote:
>
>>
>> I attempted to run ISC DHCP server 3.0.3 on a Solaris 10 container (zone)
>> but receive this error message:
>>
>> "Can't get interface flags for lo0: No such device or address"
>
>Can't you bring up a virtual lo0 in your container by issuing "ifconfig lo0 
>plumb"?
>
No. There is an alias lo0:1. Solaris zones are more like chroot on steroids than 
truely separate virtual servers. This means there are some limitastions, such as 
a zone can't see anyone else's network interfaces as local to it.

root at zdhcp1# zonename
zdhcp1

root at zdhcp1# ifconfig -a       
lo0:1: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 
index 1
        inet 127.0.0.1 netmask ff000000 
hme0:1: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
        inet 192.168.14.91 netmask ffffff00 broadcast 192.168.14.255

root at zdhcp1# ifconfig lo0 plumb
ifconfig: plumb: SIOCLIFADDIF: lo0: permission denied

I haven't tried, but it might be possible to define USE_SOCKETS and
recompile dhcpd. Alternatively not make it an error if lo0 isn't
available.

I suspect though that we won't be able to open a raw socket on the main
interface to listen on, so it's probably not going to work anyway. If
Sun is having issues, then it will probably be necessary to wait until
they sort out the kernel side of things.

regards,
-glenn



More information about the dhcp-users mailing list