Re: [mpls] Update to the bullets on the first page of my review of draft-song-mpls-extension-header-09

[Haoyu Song <[e-mail]>]

Hi Tony,

When you say "scope", do you mean network scope or in-device scope? In my second point, I mean the action scope in a device. For example, IOAM can be configured to collect data from anywhere on the processing pipeline, depending on what's set in the data type bitmap. If it's configured to collect ingress data only, then its scope become "ingress"; similarly, it's can also be configured to be "egress" only or cover the entire pipeline. This is a case that the scope is changeable and controlled by the application itself. If such kind of action has some dependency with other actions, it'll make a fairly complex situation. 

Fortunately, as a passive data collection mechanism, AFAICS IOAM can be independent with other actions. 

Then, do we have any solid case that the dependency between actions exist? I can't think of any now. If we have real examples, that would help a lot for the discussion.   

Regards,
Haoyu 

-----Original Message-----
From: Tony Li <[e-mail]> On Behalf Of Tony Li
Sent: Tuesday, September 6, 2022 10:24 AM
To: Haoyu Song <[e-mail]>
Cc: Loa Andersson <[e-mail]>; Tianran Zhou <[e-mail]>; [e-mail]; [e-mail]; [e-mail]; [e-mail]; DetNet Chairs <[e-mail]>
Subject: Re: [mpls] Update to the bullets on the first page of my review of draft-song-mpls-extension-header-09


Hi Haoyu,


> For mutually independent MNAs, enforcing the order doesn't make any sense but become an unnecessary constraint.  


You are correct.  However, we are NOT asking anyone to create a pairwise ordering between specific actions. Rather, our ask is that solutions specify the overall order of evaluation.  Further, there has been no discussion of ‘enforcement’. There are no police involved. What we are asking for is a specification.


> For some MANs, enforcing the order is also impossible (e.g., some MNA needs to be processed throughout the pipeline, and yet some others are only processed at some specific points). 


That would rather suggest separate scopes, as discussed. The order of evaluation of actions across scopes will also need to be specified.


> For dependent MNAs, their order must be specified when being introduced, but their execution order doesn't need to match their appearance order in a packet. So the processing order that jumps  back and forth from psd and isd is possible.  


Again, the order of encoding and the order of evaluation are independent. We’re asking for the specification of a deterministic ordering. A solution may choose to use order of appearance in the packet or it may not, that is up to the solution.

Tony