DPDK usage discussions
 help / color / mirror / Atom feed
From: Douglas Fischer <fischerdouglas@gmail.com>
To: users@dpdk.org
Subject: [dpdk-users] DPDK Limits by NIC - Choosing best routes
Date: Thu, 25 Feb 2021 15:20:01 -0300	[thread overview]
Message-ID: <CAKEr4RT-oJ21OoSgOnERe5Sj3a+3yjWW6CKMcTjSGxjro=TMFQ@mail.gmail.com> (raw)

 Sorry if this is a recurrent topic in this list.
I tried to search a bit on mail-archive... But I Didn't find what I guess
is relevant.


One of my major interests is using DPDK on software-based routers.
I still don't understand very well the limitations of DPDK, but I'm pretty
sure that it will mostly depend on the resources of the silicon on each
card.

Talking about Internet routing, the number of routes on FIB is a big matter.
But, allied to routing, other features like filtering/shaping/etc are also
important.

So, considering that FIB routes and Filtering resources compete by the same
limited resources on NICs.
I will also guess that would be reasonable to say that make senses do not
push all the DFZ 900K routes to a NIC.
(I'm so noobie on that that I even don't know if this would be possible.)

So, if I won't be able to push all the routes to the NIC, makes sense that
I choose the more often used routes.
(Talking about normal ITP transit packets, usually, 90% of the packets are
in between the same 20K routes.)


And now comes the question:
On the DPDK ecosystem, there is already something to deal with this choice
of the best routes?
Or it needs some extra daemon to deal with it?


An extra question:
Where can I confirm the amount of available DPDK resources on each NIC
model?


Thanks in advance!

-- 
Douglas Fernando Fischer
Engº de Controle e Automação

                 reply	other threads:[~2021-02-25 18:20 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAKEr4RT-oJ21OoSgOnERe5Sj3a+3yjWW6CKMcTjSGxjro=TMFQ@mail.gmail.com' \
    --to=fischerdouglas@gmail.com \
    --cc=users@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).