Re: same value operation attributes restriction

[Andy Bierman <abierman@cisco.com>]

At 04:17 PM 3/26/2004, Randy Presuhn wrote:
>Hi -
>
>> From: "Andy Bierman" <abierman@cisco.com>
>...
>> Sent: Friday, March 26, 2004 2:41 PM
>> Subject: RE: same value operation attributes restriction
>...
>> It's not easier on the managed device.  Rollback-on-error
>> is non-trivial to implement (that's why it's a capability).
>>
>> One of the significant reasons that SNMP failed for configuration
>> is the assumption that agents should be dumb and all the transaction
>> complexity should fall on the manager.  IMO, the agent needs to
>> provide appropriate transaction capability, even if that's not easy.
>...
>
>I'm missing something in the big-picture logic here.
>
>One of the objections to SNMP was that the logic in the managed system
>needed for commit/rollback/undo was too complex for implementors to get
>right.  The netconf argument runs that this should be a capability and that market
>pressure will determine whether it is widely implemented.  What escapes
>me is how the market pressure on implementors to do support rollback-on-error
>will be any different for netconf (where it is optional) than for SNMP (where it
>is mandatory).

SNMP allows an arbitrary mix of arbitrarily incomplete high-level
operations. NETCONF allows an arbitrary mix of complete high-level
operations.  This is easier to implement.  I also said this was one 
of the factors, not the only one.  Vendors will implement this if
there is enough customer demand.  


>Randy

Andy


--
to unsubscribe send a message to netconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/netconf/>