[Ndn-interest] sending queue

[Junxiao Shi]

Hi Jinyang

The correct mailing list to ask NFD questions is *nfd-dev*, not
ndn-interest. Please send future questions to nfd-dev instead.

NFD does not explicitly maintain queues. *Both receive queue and send queue
reside in Linux or macOS kernel*, in the form of socket buffers. You
may *change
queue capacity by configuring kernel socket buffer sizes*, including UDP,
TCP, and libpcap.
Subclasses of ‘Transport’ pulls packets from the receive queue and places
packets into the send queue via Boost.Asio library. *Transport queries
socket API to retrieve send queue length and send queue capacity*, and
GenericLinkService inserts congestion marks when it detects send queue
filling up. Transport does not consider receive queue length and receive
queue capacity.

Yours, Junxiao

On Fri, Jun 8, 2018 at 22:06 jinyangl at mail.ustc.edu.cn <
jinyangl at mail.ustc.edu.cn> wrote:

> Hello,
> I encountered a problem When sending a lot of interest packets. Is there
> a sending queue to store interest packets?
> If so, is it implemented in NFD or is it implemented by a router that is
> independent of NFD?
> ------------------------------
> jinyangl at mail.ustc.edu.cn
> _______________________________________________
> Ndn-interest mailing list
> Ndn-interest at lists.cs.ucla.edu
> http://www.lists.cs.ucla.edu/mailman/listinfo/ndn-interest
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndn-interest/attachments/20180613/6f1ff06d/attachment.html>