All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Ethy H. Brito" <ethy.brito@inexo.com.br>
To: "Toke Høiland-Jørgensen" <toke@redhat.com>
Cc: xdp-newbies@vger.kernel.org, Jesper Brouer <brouer@redhat.com>
Subject: Re: xdp-cpumap-tc multi queue schema and traffic mirroring
Date: Fri, 8 Oct 2021 10:50:50 -0300	[thread overview]
Message-ID: <20211008105050.5c70c225@pulsar> (raw)
In-Reply-To: <875yu8n4xp.fsf@toke.dk>

On Thu, 07 Oct 2021 19:02:58 +0200
Toke Høiland-Jørgensen <toke@redhat.com> wrote:

> "Ethy H. Brito" <ethy.brito@inexo.com.br> writes:
> 
> > Hi All
> >
> > Before I moved to XDP cpumap bandwidth control "philosophy" I used to snoop
> > inside the htb classes (mainly the default one) mirroring traffic to a
> > dummy interface then run tcpdump on that like:
> >
> > 	tc filter add dev eth0 parent 1: protocol all prio 0xffff \
> > 	u32 match u32 0 0 flowid 1:$shp action mirred egress mirror dev
> > dummy0
> >
> > Then "tcpdump -npi dummy0" used to show me all that classid "1:$shp" was
> > carrying.
> >
> > Now, with the multi queue environment XDP creates, I can't make mirroring
> > to work. For instance: I need to see what is going on with HTB classid
> > e:102, so I tried: 
> > 	tc filter add dev eth0 parent e: ... e:102 action mirred egress ...
> >
> > and "tcpdump -npi dummy0" prints traffic that has nothing to do with
> > classid e:102. I can confirm the traffic is pored thru the class since its
> > bandwidth is controlled/limited correctly.  


Hi Toke, et All.

> 
> The xdp-cpumap-tc utility will attach a filter to the egress interface,
> presumably that is taking priority and short-circuit your mirred filter?

This makes sense. 
But How to circumvent this??

Snooping the default class (classes, in xdp-cpumap case) is essencial to catch misconfigured mappings.

Regards

Ethy

  reply	other threads:[~2021-10-08 13:51 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-06 21:18 xdp-cpumap-tc multi queue schema and traffic mirroring Ethy H. Brito
2021-10-07 17:02 ` Toke Høiland-Jørgensen
2021-10-08 13:50   ` Ethy H. Brito [this message]
2021-10-08 17:36     ` Toke Høiland-Jørgensen

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=20211008105050.5c70c225@pulsar \
    --to=ethy.brito@inexo.com.br \
    --cc=brouer@redhat.com \
    --cc=toke@redhat.com \
    --cc=xdp-newbies@vger.kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.