From: Thea Corinne Rossman <thea.rossman@cs.stanford.edu>
To: users@dpdk.org
Subject: DPDK with Containernet
Date: Fri, 22 Nov 2024 11:51:04 -0800 [thread overview]
Message-ID: <CAHXF6-pWLsT4gfj2YLvE3V_O9k1D-NrUkBweJvttMqz66zJseA@mail.gmail.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 1724 bytes --]
Hello! Has anyone here used Containernet with DPDK before? Our use-case is
evaluating DPDK-based applications under different network conditions
(e.g., lossy links).
I've watched this talk https://www.youtube.com/watch?v=KPshkuEPOOU, and
I've done pretty extensive research into running DPDK applications in
containers for intra-host container networking. (+ thank you to all those
who responded to my previous question on this forum!)
My understanding is that there are two approaches to DPDK + container
networking:
- Aggregation: Using VirtIo-User (containers) and vhost-user (backend)
which offers a software PMD. This also needs to be connected to a virtual
switching infrastructure on the host.
- Slicing: Using SR-IOV or similar to give containers direct access to
devices.
I'm not sure how to reconcile these options with Containernet, both
conceptually and in practice.
Conceptually:
- Does Containernet's OVS
<https://github.com/containernet/containernet/blob/8e0aabef8f23def3961b30897dd486191d9bd693/util/install.sh#L406>/switching
provide
the appropriate infrastructure for the aggregation option?
- Is Containernet even compatible with the VF/slicing option? Can I just
build a Containernet node from a DPDK-compatible Dockerfile, set it up with
the correct (sliced) PCI addresses, and then connect it to Containernet's
emulated network? Would I have to configure anything on the emulated
network to make this work?
In practice: this seems like a really common use-case (DPDK is specifically
mentioned in the original Containernet talk!). However, I can't find any
clear guides or images for this, in the Containernet repo or otherwise.
Thank you in advance!
Thea
[-- Attachment #2: Type: text/html, Size: 2029 bytes --]
reply other threads:[~2024-11-22 19:51 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=CAHXF6-pWLsT4gfj2YLvE3V_O9k1D-NrUkBweJvttMqz66zJseA@mail.gmail.com \
--to=thea.rossman@cs.stanford.edu \
--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).