Hi Greg,

I agree that network slicing would mainly be used for URLLC services, which 
have stringent requirement on performance. To meet such requirement, we need to 
consider the E2E transport paths. FlexE can be one candidate technology for 
interface level channelization and isolation, and we also need to consider 
other pieces in the end-to-end paths.

OTOH, as mentioned in the enhanced VPN presentation, the mechanism used to 
represent the underlay resources can be general and independent from the 
implementation of the resource partitioning, such as using SR SIDs to represent 
per-hop resource allocation.

Best regards,
Jie

From: rtgwg [mailto:[email protected]] On Behalf Of Greg Mirsky
Sent: Thursday, November 16, 2017 2:53 PM
To: [email protected]
Subject: Scope of strict SLA in network slicing

Dear All,
I'd like to stress what I've said at the mike. Two of three network slicing 
scenarios, enhanced mobile broadband and massive IoT, that have been defined by 
3GPP do not set forth too strict latency/jitter/packet loss constraints. But 
netslices to support ultra-reliable and low latency communications (URLLC) are 
expected to have very stringent latency/jitter/packet loss constraints e2e. 
But, in my opinion, there's no requirement that URLLC slices must share 
physical resources. One of solutions may use, for example, FlexE to deliver CBR 
guaranteed services.

Regards,
Greg
_______________________________________________
rtgwg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/rtgwg

Reply via email to