[Nfd-dev] [EXT] ndnd memory increasing

[Junxiao Shi]

Hi Hao

A quick search of "HTTP over NDN" finds this:
Y. Huang, J. Luo, C. Mo and L. Cui, "HTTP/3 over Information-Centric
Networking," 2024 IEEE/ACM 32nd International Symposium on Quality of
Service (IWQoS), Guangzhou, China, 2024, pp. 1-2, doi:
10.1109/IWQoS61813.2024.10682843. keywords: {Transport
protocols;Information-centric networking;Quality of
service;Browsers;Security;System analysis and design},

Nevertheless, it does not feel beneficial to run HTTP/3 over NDN.
To fully take advantage of NDN benefits (especially the security benefits),
the applications have to be redesigned to use NDN semantics, and there's no
way around it.

Yours, Junxiao

On Mon, Jun 23, 2025 at 7:46 AM 郝广宇 <haoguangyu at chorustek.com> wrote:

>
> *External Email*
> ------------------------------
> Hi Lixia
>
> 1. Is there any solution for HTTP/3 over NDN?
> 2. It's difficult for us to abandon the currently used HTTP protocol and
> adopt end-to-end NDN. How have you considered this aspect?
>
> thanks
> ------------------ Original ------------------
> *From: * "郝广宇"<haoguangyu at chorustek.com>;
> *Date: * Mon, Jun 23, 2025 08:26 AM
> *To: * "Lixia Zhang"<lixia at cs.ucla.edu>;
> *Cc: * "shijunxiao"<shijunxiao at arizona.edu>;
> *Subject: * Re: [Nfd-dev] [EXT] ndnd memory increasing
>
> Hi Lixia
>
> This does indeed lead to a messy protocol stack. We have already
> implemented HTTP/1.1 over NDN, but we are facing issues with loading
> high-resolution videos. We are considering using HTTP/3 over NDN to address
> the problem of high-resolution video playback.
>  thanks
>
> ------------------ Original ------------------
> *From: * "Lixia Zhang"<lixia at cs.ucla.edu>;
> *Date: * Mon, Jun 23, 2025 06:47 AM
> *To: * "郝广宇"<haoguangyu at chorustek.com>;
> *Cc: * "shijunxiao"<shijunxiao at arizona.edu>;
> *Subject: * Re: [Nfd-dev] [EXT] ndnd memory increasing
>
>
> On Jun 20, 2025, at 4:56 PM, 郝广宇 <haoguangyu at chorustek.com> wrote:
>
> Hi Junxiao
>
> Can only NDN over HTTP/3 be implemented? We are hoping to achieve HTTP/3
> over NDN — is there any solution for that?
>
> thanks
>
> I do not know the specific reason of why you want to run NDN over HTTP/3,
> just FYI IETF masque working group is developing solutions to "allow
> configuring and concurrently running multiple proxied stream- and
> datagram-based flows inside an HTTP connection."
> In particular, see internet-draft
> https://datatracker.ietf.org/doc/draft-ietf-masque-connect-udp-listen/
> <https://datatracker.ietf.org/doc/draft-ietf-masque-connect-udp-listen/>
>
> NDN can run over UDP tunnels, so in theory one could do NDN-UDP-HTTP,
> although that makes an ugly protocol stack (NDN on UDP, UDN on HTTP3,
> HTTP/3 on QUIC, and QUIC on UDP again).
>
> Lixia
>
> ------------------ Original ------------------
> *From: * "Junxiao Shi"<shijunxiao at arizona.edu>;
> *Date: * Fri, Jun 20, 2025 07:57 PM
> *To: * "郝广宇"<haoguangyu at chorustek.com>;
> *Cc: * "nfd-dev"<nfd-dev at lists.cs.ucla.edu>;
> *Subject: * Re: [EXT] [Nfd-dev] ndnd memory increasing
>
> Hi Hao
>
> HTTP/3 WebTransport in ndnd only has the server side.
> The client side is in NDNts library and runs in web browsers.
>
> To establish communication between two ndnd instances, use the UDP
> transport instead.
>
> Yours, Junxiao
>
> On Fri, Jun 20, 2025, 04:57 郝广宇 <haoguangyu at chorustek.com> wrote:
>
>>
>> *External Email*
>> ------------------------------
>> Hi Junxiao
>>
>> If an HTTP/3.0 client can connect to ndnd, how can an HTTP/3.0 server
>> connect to ndnd?
>>  thanks
>>
> _______________________________________________
> Nfd-dev mailing list
> Nfd-dev at lists.cs.ucla.edu
> https://www.lists.cs.ucla.edu/mailman/listinfo/nfd-dev
> <https://www.lists.cs.ucla.edu/mailman/listinfo/nfd-dev>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/nfd-dev/attachments/20250625/817e0a71/attachment-0001.htm>