Hi Lizhong,
On 10/06/2009 03:58, Lizhong Li wrote:
> Anurag,
>
> I found the ouptput of nwamadm was changed, I'm curious why it's 'offline*',
> what's the meaning of '*' ? Also should the original unplumbed link be plumbed
> when the nwamd starts ? It's not reasonable somewhat.
>
> earthscience:nwam# nwamadm list
> TYPE PROFILE STATE
> ncp Automatic online
> ncu:phys nge1 online
> ncu:ip nge1 offline*
> ncu:phys e1000g0 online
> ncu:ip e1000g0 offline*
> ncu:phys e1000g1 online
> ncu:ip e1000g1 offline*
> ncu:phys nge0 online
> ncu:ip nge0 online
> loc NoNet offline
> loc Automatic online
> loc Legacy offline
>
When it's in the offline* state it mean that it hasn't been able to be brought
on line for some reason. That reason is an auxiliary state, which is able to be
seen if you provide the "undocumented" (intentionally AFAIK) -x option.
Since the change in the state machine, any i/f isn't deemed online until it
really is fully available - in the case you've got, it's quite likely that it's
in the state "waiting for dhcp" or similar - I've certainly experienced a bug
where an address was though to be un-assigned by nwamd, yet there was one with
ifconfig, like you have....
You can find out more about the use of these new states at the updated docs at:
http://www.opensolaris.org/os/project/nwam/p1spec/daemon/
As for whether i/f's are plumbed or not - it's my understanding that if the phys
ncu isn't online (e.g. a lower prio group is active) then it won't be plumbed -
this has caused some issues with some wifi drivers which assume that the i/f
should be plumbed before a wifi-connect will work - the nwam team consider this
a bug in the driver, and were going to contact the driver team about this.
HTH,
Darren.
