DPDK patches and discussions
 help / color / mirror / Atom feed
From: Aman Thakur <r.aman.t.435@gmail.com>
To: "users@dpdk.org" <users@dpdk.org>
Cc: dev@dpdk.org
Subject: Incorrect link speed for DPDK Bond in Balance RR Mode
Date: Mon, 24 Feb 2025 18:55:50 +0530	[thread overview]
Message-ID: <CAOUH2--EMs+S54b4JyKRz1=_msbLzUn=oaNGVGF0ptjMa-eM1Q@mail.gmail.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 2012 bytes --]

Dear Sir/ma'am,

I've created a DPDK bond with two slave interfaces, each capable of 1Gbps
(mtu 1500 and both Full duplex). I'm using bonding mode 0 (balance RR -
round-robin). Theoretically, this should give me a 2Gbps link. However, the
bond interface only reports a speed of 1Gbps, and I've noticed that one of
the slave interfaces is never active. Here's a breakdown of my setup and
the issue:

Here's a breakdown of my setup and the issue:

   - *DPDK Version:* 21.11.9
   - *Bonding Mode:* 0 (balance RR)
   - *Number of Slaves:* 2
   - *Slave Speed:* 1Gbps each, same MTU and each operating at full duplex.
   - *Expected Bond Speed:* 2Gbps
   - *Observed Bond Speed:* 1Gbps
   - *Problem:* One slave interface remains inactive. Both slaves were
   active *before* calling rte_eth_dev_start(BOND_PORT).
   - *Lcore Usage:* Currently using 1 lcore. Tried with 2 and 3 lcores, but
   the problem remains.
   - Output of *rte_eth_bond_active_slaves_get *gives 1 implying only 1
   slave is active at a time.

I'm following the DPDK bonding example (specifically the port
initialization) found here:
https://doc.dpdk.org/api-17.08/examples_2bond_2main_8c-example.html, ( for
reference only I've used sample from 21.11.9 only )

*My Questions:*

   1. What could be causing one of the slaves to become inactive after
   rte_eth_dev_start(BOND_PORT)?
   2. Is there a specific configuration or step I might be missing that's
   preventing the bond from utilizing both slaves in round-robin mode?
   3. How can I troubleshoot this issue to determine why the second slave
   is not participating in the bond? Are there specific DPDK tools or logs I
   should be examining?
   4. Are there any known compatibility issues or limitations with balance
   RR mode that could explain this behavior?
   5. Could the number of lcores be related to this issue, even though I've
   tried multiple values? Is there a minimum or recommended number of lcores
   for bond mode 0 with 2 slaves?

Thanks and Regards

Aman Thakur

[-- Attachment #2: Type: text/html, Size: 2489 bytes --]

                 reply	other threads:[~2025-02-25  8:09 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='CAOUH2--EMs+S54b4JyKRz1=_msbLzUn=oaNGVGF0ptjMa-eM1Q@mail.gmail.com' \
    --to=r.aman.t.435@gmail.com \
    --cc=dev@dpdk.org \
    --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).