DPDK usage discussions
 help / color / mirror / Atom feed
From: piyush gupta <pgupta9393@gmail.com>
To: users@dpdk.org
Subject: Performance issues while using DPDK with azure accelerated networking
Date: Fri, 23 Dec 2022 12:07:59 +0530	[thread overview]
Message-ID: <CANSLT0aKvoNB0EX6c_0Feh1TcM_SoUBj6hgzytgoFt=OWBa6aw@mail.gmail.com> (raw)


[-- Attachment #1.1: Type: text/plain, Size: 1217 bytes --]

Hi,

I am running a *D8sV3* type instance on azure with *RHEL7.9* & accelerated
networking enabled on one of the interface.

On this am running my DPDK app which is binding to the VF exposed & its
corresponding netvsc interface via *vdev_netvsc driver* as suggested in
https://learn.microsoft.com/en-us/azure/virtual-network/setup-dpdk
<https://learn.microsoft.com/answers/storage/attachments/267669-dpdk-cpus-perf.png>

DPDK doc : vdev_netvsc.html
<https://learn.microsoft.com/answers/storage/attachments/267669-dpdk-cpus-perf.png>

The expected behaviour is that all traffic should flow through the physical
interface (VF) and not from synthetic device (eth2 in my case).
But I can see all traffic coming to eth2 & dtap0 devices. This is causing
my dpdk cpus to handle RX interrupts for these packets causing major
performance issues & jitters. I am not understanding why all these packets
are being mirrored to slow path.

*Note: All my traffic is UDP.*


[image: dpdk_cpus_perf.png]

[image: azure_pkts_on_eth2.png]

I would like to know if there is any setting I am missing here. This is
causing a lot of problems for me to qualify on Azure cloud with expected
throughput.


Thanks.

Best regards,

Piyush Gupta

[-- Attachment #1.2: Type: text/html, Size: 5036 bytes --]

[-- Attachment #2: dpdk_cpus_perf.png --]
[-- Type: image/png, Size: 57553 bytes --]

[-- Attachment #3: azure_pkts_on_eth2.png --]
[-- Type: image/png, Size: 48595 bytes --]

             reply	other threads:[~2022-12-24 10:02 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-23  6:37 piyush gupta [this message]
2022-12-24 11:16 ` Hui Ling

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='CANSLT0aKvoNB0EX6c_0Feh1TcM_SoUBj6hgzytgoFt=OWBa6aw@mail.gmail.com' \
    --to=pgupta9393@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).