[{"author": "Dhruv Dhody", "text": "<p>Help Andrew with notes at <a href=\"https://notes.ietf.org/notes-ietf-121-pce?edit\">https://notes.ietf.org/notes-ietf-121-pce?edit</a></p>", "time": "2024-11-06T15:00:14Z"}, {"author": "John Scudder", "text": "<p>\ud83d\udc4b\ud83c\udffb</p>", "time": "2024-11-06T15:02:10Z"}, {"author": "John Scudder", "text": "<p>As if an hour ago the allocation policy draft is back to waiting on authors. :-)</p>", "time": "2024-11-06T15:06:54Z"}, {"author": "John Scudder", "text": "<p>(As of)</p>", "time": "2024-11-06T15:07:02Z"}, {"author": "John Scudder", "text": "<p>Also thanks for the prod on the early alloc renewals. I found two pending, approved. If I missed any please LMK.</p>", "time": "2024-11-06T15:08:41Z"}, {"author": "Cheng Li", "text": "<p>authors(state-sync) will update the draft accordingly after the meeting</p>", "time": "2024-11-06T15:11:10Z"}, {"author": "Quan Xiong", "text": "<p>Thanks, Dhruv. Authors will add the implementation section in draft-ietf-pce-entropy label</p>", "time": "2024-11-06T15:12:52Z"}, {"author": "Julien Meuric", "text": "<p>@John: Thanks a lot for the renewal approvals. <span aria-label=\"+1\" class=\"emoji emoji-1f44d\" role=\"img\" title=\"+1\">:+1:</span></p>", "time": "2024-11-06T15:20:28Z"}, {"author": "Samuel Sidor", "text": "<p>I was just describing that it is more about spring CS draft depending on extensions PCEP CS draft  (even if that is also not strong dependency)) rather than PCEP draft depending on spring draft.</p>", "time": "2024-11-06T15:26:48Z"}, {"author": "Dhruv Dhody", "text": "<p>best to progress PCE, IDR and SPRING together</p>", "time": "2024-11-06T15:27:32Z"}, {"author": "Samuel Sidor", "text": "<p>I just hoped that we can proceed with PCEP draft even sooner as extensions are potentially re-usable for other use-cases as well (not just for Circuit style policies), only \"connection\" to SPRING draft is just that SPRING draft is explaining motivation behind preparing those extensions.</p>", "time": "2024-11-06T15:32:16Z"}, {"author": "Dhruv Dhody", "text": "<p>good point</p>", "time": "2024-11-06T15:32:38Z"}, {"author": "Boris Khasanov", "text": "<p>@Samuel, exactly, I even wanted to ask -can we apply those extensions for  other cases</p>", "time": "2024-11-06T15:33:40Z"}, {"author": "Samuel Sidor", "text": "<p>@Boris - yes, it should be re-usable even for other path setup types.</p>", "time": "2024-11-06T15:35:04Z"}, {"author": "Boris Khasanov", "text": "<p>@Samuel -cool!</p>", "time": "2024-11-06T15:35:35Z"}, {"author": "John Scudder", "text": "<p>For what it's worth (possibly not much :-) this sounds very similar to the dynamic-capability proposal for BGP, including the \"maybe it will be very hard to do the differential change and then just reset\" part. As a result dyn-cap never has made progress.</p>", "time": "2024-11-06T15:45:54Z"}, {"author": "John Scudder", "text": "<p>Oh ha ha Ketan is saying similar.</p>", "time": "2024-11-06T15:46:10Z"}, {"author": "Julien Meuric", "text": "<p>We have consensus among BGP experts. ;-)</p>", "time": "2024-11-06T15:46:36Z"}, {"author": "John Scudder", "text": "<p>I lost audio. Sort of sounds like Stephane's point is similar, though, to what I was saying about \"make resetting a session something you aren't afraid to do\".</p>", "time": "2024-11-06T15:54:38Z"}, {"author": "Dhruv Dhody", "text": "<p>something like graceful restart</p>", "time": "2024-11-06T15:55:29Z"}, {"author": "John Scudder", "text": "<p>Precisely.</p>", "time": "2024-11-06T15:56:23Z"}, {"author": "John Scudder", "text": "<p>Unfortunately I think you will find that while the problems aren't quite isomorphic they are at least related -- the more you try to make the restart low-impact, the more effort you have to expend ensuring your solution is correct. Differently put, the more you expand the bug surface. :-( </p>\n<p>\"You can pay me now or you can pay me later.\" <br>\n\"Good, fast, cheap, pick two.\"</p>", "time": "2024-11-06T15:58:40Z"}, {"author": "Andrew Stone", "text": "<p>thanks for the BGP comparisons and pointers -&gt; agreed, definitely worth exploring more on what happened within BGP and see applicability here with open refresh, and/or alternative -&gt; make session flap less painful</p>", "time": "2024-11-06T16:08:42Z"}, {"author": "Samuel Sidor", "text": "<p>@Andrew: Just for reference (that BGP draft with dynamic capabilities): <a href=\"https://datatracker.ietf.org/doc/draft-ietf-idr-dynamic-cap/\">https://datatracker.ietf.org/doc/draft-ietf-idr-dynamic-cap/</a></p>", "time": "2024-11-06T16:14:29Z"}, {"author": "Andrew Stone", "text": "<p>-00  -- 2001. </p>\n<p>oh my</p>", "time": "2024-11-06T16:15:50Z"}, {"author": "Andrew Stone", "text": "<p>ty for the link though :)</p>", "time": "2024-11-06T16:15:57Z"}, {"author": "Andrew Stone", "text": "<p>(and will give it a read)</p>", "time": "2024-11-06T16:16:37Z"}, {"author": "John Scudder", "text": "<p>X-over-QUIC seems to be having a moment. Hopefully the Q/A will drill down on whether the motivation solves problems that are genuinely... problems, for PCE. (E.g. the \"session mobility\" motivation doesn't strike me as germane.)</p>", "time": "2024-11-06T16:25:29Z"}, {"author": "John Scudder", "text": "<p>And it sounded to me like Andrew just said HOL blocking isn't really a thing.</p>", "time": "2024-11-06T16:26:32Z"}, {"author": "Andrew Stone", "text": "<p>yes. HOL over the TCP channel is not typically the concern. HOL of -actually- computing paths or managing state (after transmission) is more of a concern and an implementation concern</p>", "time": "2024-11-06T16:28:39Z"}]