5 ms·
I'd love to hear more about the difference between QoS and "prioritization", as they are calling it.
by apeace 9y ago
I'd love to hear more about the difference between QoS and "prioritization", as they are calling it.
- pdabbadabba 9y agoThe existing rules contain an exception to the no-throttling and no-blocking rules for "reasonable network management." Here is how the Commission has defined that term: > A network management practice is a practice that has a primarily technical network management justification, but does not include other business practices. A network management practice is reasonable if it is primarily used for and tailored to achieving a legitimate network management purpose, taking into account the particular network architecture and technology of the broadband Internet access service. It's debatable what exactly this means, but providing higher QoS for traffic that needs it--in a way that does not discriminate between providers of those services--would seem to be a strong candidate.
- rosser 9y ago> in a way that does not discriminate between providers of those service Exactly. Anti-NN (broadly) wants to be able to discriminate on the basis of provider, not service.
- pishpash 9y agoI take "legitimate network management purpose" to mean improving some metric of overall usage satisfaction. This is tricky, and already leaves a lot of room to maneuver.
- nitrogen 9y agoQoS is decided by the user/service. Paid prioritization is decided by the ISP. Every connection can have a max high-priority bandwidth and a max best-effort bandwidth, so people won't benefit from just marking every connection as high-QoS.
- dfox 9y agoAny sane ISP either ignores or replaces QoS tags on ingress to their network.
- wmf 9y agoCurrently, ISPs strip/normalize the TOS bits on incoming traffic. ISPs could enable DiffServ (especially at the CMTS/DSLAM) and charge content providers to respect their TOS bits, allowing traffic like VoIP and gaming to get lower latency/jitter. But they don't do that. Instead, ISPs are creating artificial congestion on their transit links (by leaving router ports unplugged) and then selling that bandwidth to content providers via paid peering to avoid the congestion. Basically "it would be a shame if massive packet loss were to happen... never mind, we already did it". So yeah, paid peering probably does have lower latency than congested and bufferbloated transit links, but it's not like ISPs are using any kind of QoS as the term is traditionally understood. ISPs always talk about low-latency applications (think of the deaf people!) but then they don't do anything that would actually reduce latency/jitter.