FYI on two related items of interest to this list.

1 – If you are a Comcast customer, consider volunteering for our upcoming low latency networking trials (using IETF L4S). See my blog post at https://corporate.comcast.com/stories/comcast-kicks-off-industrys-first-low-latency-docsis-field-trials and signup at https://www.xfinityinsightscommunity.com/register/start/f4f1ea42-247c-44f1-bea0-c2d7178cfe00.

2 – The IETF L4S standard and the upcoming NQB standards – both dual-queue low latency networking – will rely on end-to-end packet marking to signal dual-queue-capable routers to send packets to the low latency queue. Thus, there will be inter-domain marking which will affect many networks’ edge configurations.

• In the case of L4S it may be easier, as L4S relies on the ECN header – and most networks probably forward packets without touching this header. The L4S setting here is to set ECT(1) or CE.
• In the case of NQB, the marking is expected to use DSCP-45 (once the RFCs are issued, IANA will make the code point assignment). This will be more challenging for those networks that scrub any DSCP marks from packets and over-write with their own. Post-standardization, networks should permit inbound DSCP-45 marks to remain & pass them along down to the user edge (in cases where 45 is used internally, networks may remark internally and then mark back to 45 as it goes to the user LAN).
• In both cases – ECN or DSCP – the packets remain at the best effort priority level.
• For slightly more detail see https://github.com/jlivingood/IETF-L4S-Deployment/blob/main/Network-Config-Guide.md

Thanks!
Jason