Marco,

Thanks for explaining your rationale, this helps clarify my thinking.  As you 
noticed, my bias was leading me to a certain interpretation of the meaning of 
the proposed fields.  I suspect that there will be use cases of interest where, 
as you mentioned in your next to last paragraph, clients of carrier 1 need to 
connect to an agent of carrier 2.   In any event, I am interested in continued 
discussion on this topic.

Best regards,

Larry

From: Marco Liebsch [mailto:[email protected]]
Sent: Friday, May 29, 2015 8:34 AM
To: Laporte, Laurent [CTO]; [email protected]
Subject: RE: FPC carrier ID and network ID question

Hi Larry,

thanks for your feedback and valid doubts, as you pointed out an important 
point to discuss.
First of all, the space for a certain field in the ID can be increased. We may 
have a preceding
discussion about how the identifier format should look like; as proposed in 
this version of
the draft or differently.

The rationale behind the Client-ID, Agent-ID and DPN-ID to allow unambiguous 
identification
of a function instance associated with FPC and its location. 'Location' means 
to identify the
network, e.g. a certain datacenter, where a function is instantiated and 
operational.

Example: An IP switch, which can serve as Data Plane Node for mobility 
management, is
located in a local POP/datacenter. The datacenter can be identified in the 
Network ID field of
the complete identifier and should be unique within the carrier's network 
topology. In the example
you brought, the Network Code (MNC) identifies rather the carrier instead of a 
certain
spot of a single carrier's network topology.

In terms of FPC deployment, we may omit the Carrier-ID field in identifiers in 
case
we do not expose and use these identifiers outside of a single carrier. But it 
may be
useful to identify the carrier as well in case, for example, a Client of 
carrier 1 connects to an Agent
of Carrier 2. In case we need to keep a Carrier-ID field, it may comprise the 
complete
tuple of MCC/MNC as you refer to.

I hope that clarifies your question and we can follow up on that thread to find 
a suitable
format for the complete identifier.

Thanks,
Marco



From: dmm [mailto:[email protected]] On Behalf Of Laporte, Laurent [CTO]
Sent: Donnerstag, 28. Mai 2015 18:33
To: [email protected]<mailto:[email protected]>
Subject: [DMM] FPC carrier ID and network ID question

Hello,

I have a question regarding the Carrier ID and Network ID elements proposed in 
the DMM FPC protocol.  I am trying to  a) understand the meaning intended for 
Carrier and Network; and, b) reconcile these names with what is familiar to me 
as someone who works for a mobile service provider.  The problem that I'm 
basically running into is that there are only 8 bits assigned each for Carrier 
ID and Network ID.  I am accustomed to utilizing a PLMN, which is composed of 
an MCC and MNC, each having allowed values up to 999.  With only eight bits, 
there are fewer values than required for either MCC or MNC.  Eight bits for 
Carrier ID and Network ID seems to be quite constrained to me.

I suspect that something else is meant, but I cannot readily surmise what it 
might be.

Thanks,

Larry Laporte


________________________________

This e-mail may contain Sprint proprietary information intended for the sole 
use of the recipient(s). Any use by others is prohibited. If you are not the 
intended recipient, please contact the sender and delete all copies of the 
message.

________________________________

This e-mail may contain Sprint proprietary information intended for the sole 
use of the recipient(s). Any use by others is prohibited. If you are not the 
intended recipient, please contact the sender and delete all copies of the 
message.
_______________________________________________
dmm mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmm

Reply via email to