Sangeeta Misra wrote:
> On 03/02/09 20:49, michael schuster wrote:
>> Kacheong Poon wrote:
>>> Michael Schuster wrote:
>>>> On 02/25/09 15:46, Michael Schuster wrote:
>>>>
>>>>> since there's been no other comments, I'll proceed as follows:
>>>>>
>>>>> - remove the "serverID" keyword again.
>>>>> - since the base of the serverID is now the SG name, remove the 
>>>>> ability to use the hostname as (base) serverID.
>>>>> - (this is according to the initial plan) modify commands which 
>>>>> used to manipulate servers by address to now handle server IDs.
>>>>
>>>> Would it be better to not only change the commands to accept server 
>>>> IDs instead of IP addresses, but also change the *keyword* from 
>>>> "server" to something else, so there's no (well ... less) confusion: 
>>>> so instead of:
>>>
>>>
>>> Isn't the serverID (base + a number) assigned by the system?
>>> And the number part may possibly change when the same server
>>> is removed and then added back.  Correct?  If the above are
>>> true, then an admin may not know the "current" serverID and
>>> it means that requiring a serverID to be specified can be
>>> problematic.
>>
>> ahem ... the whole notion of the server ID was created to have 
>> something other than the ip address to manipulate servers with. What 
>> do you suggested we do?
>>
>> Michael
> Michael,
> What do you require the user to specify in server id (  basename, 
> basename + number). I thought at one point you had decided that the 
> basename and the number will be asigned by the system, with the basename 
> being the server group name( and that the mapping of serverid and IP 
> address would be listed in list-servergreoup command).  Have you changed 
> your mind on that ?

no - I didn't notice that being said or implied.

Michael
-- 
Michael Schuster     http://blogs.sun.com/recursion
Recursion, n.: see 'Recursion'

Reply via email to