DPDK usage discussions
 help / color / mirror / Atom feed
From: Hong Christian <hongguochun@hotmail.com>
To: "users@dpdk.org" <users@dpdk.org>,
	Viktor Vasiliev <vasiliev-viktor@protei-lab.ru>
Cc: Yicai Lu <luyicai@huawei.com>,
	Krzysztof Kanas <kkanas@marvell.com>,
	Declan Doherty <declan.doherty@intel.com>
Subject: bond mode 4 lacp system id change cause link lost
Date: Mon, 27 Feb 2023 08:46:42 +0000	[thread overview]
Message-ID: <SY4P282MB2758E1B77E53A7DA0BF4C36AACAF9@SY4P282MB2758.AUSP282.PROD.OUTLOOK.COM> (raw)
In-Reply-To: <SY4P282MB27589ACB320E459BA1520828ACAF9@SY4P282MB2758.AUSP282.PROD.OUTLOOK.COM>

[-- Attachment #1: Type: text/plain, Size: 1388 bytes --]

Hi DPDK users,

I met a bond issue with dpdk bond mode 4.

ENV: DPDK 20.11.7
EAL parameters: --file-prefix node-1 --log-level pmd.net.bond:8 --vdev net_bonding0,mode=4,slave=d8:00.0,slave=d9:00.0 --vdev net_bonding1,mode=4,slave=d8:00.1,slave=d9:00.1

Here we only forcus on net_bonding0, which has two slave ports(1,3) in this bonding, and I set a bond mac 11:22:33:44:55:66 in app.
The bond link is lost while shutdown one slave port. The reason seems the actor system id is changed while shutdown port 0. It cause port 3 to renegotiate and link down in this senario.
I compared this with a linux bond behavior, the difference is that linux bonding device do not change system id while I shutdown one port.

Do you have any idea for the issue ? Is this a dpdk bug ?

I search the email list, and I found @Viktor Vasiliev<mailto:vasiliev-viktor@protei-lab.ru> met a slimilar issue "bond mode 8023ad LACP system id changes when link goes down<http://patchwork.dpdk.org/project/dpdk/patch/20171228021231.15363-1-3chas3@gmail.com/>", I use your patch and it fix bonding status issue but no user pkts could send out while I shutdown one port ... so it's still not valid for me. Do you have any progress on your issue ? Thanks in advance.

Sorry to cc you, dpdk experts, I notice you all fix some bond related issues, may you could help, thanks you so much!

Br,
Christian

[-- Attachment #2: Type: text/html, Size: 5387 bytes --]

           reply	other threads:[~2023-02-27  8:46 UTC|newest]

Thread overview: expand[flat|nested]  mbox.gz  Atom feed
 [parent not found: <SY4P282MB27589ACB320E459BA1520828ACAF9@SY4P282MB2758.AUSP282.PROD.OUTLOOK.COM>]

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=SY4P282MB2758E1B77E53A7DA0BF4C36AACAF9@SY4P282MB2758.AUSP282.PROD.OUTLOOK.COM \
    --to=hongguochun@hotmail.com \
    --cc=declan.doherty@intel.com \
    --cc=kkanas@marvell.com \
    --cc=luyicai@huawei.com \
    --cc=users@dpdk.org \
    --cc=vasiliev-viktor@protei-lab.ru \
    /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).