All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Brian J. Murrell" <brian@interlinx.bc.ca>
To: lartc@vger.kernel.org
Subject: Re: [LARTC] Routing problem (RTNETLINK answers: Invalid argument)
Date: Sun, 18 Feb 2007 20:30:33 +0000	[thread overview]
Message-ID: <1171830634.7053.58.camel@pc.ilinx> (raw)
In-Reply-To: <200702131450.13852.paul@diasoft.nl>


[-- Attachment #1.1: Type: text/plain, Size: 2003 bytes --]

On Thu, 2007-02-15 at 00:30 +0100, Paul Viney wrote:
> Wow! That made a difference. One 
> echo "0" > /proc/sys/net/ipv4/conf/eth1/rp_filter
> and everything started working. 
> Thanks a lot Torsten and Alex - I wouldn't have solved it without your 
> suggestions.

But the question I have had about this subject is when one has two
default routes, load balanced, does this evaluation of which interface
would be used when the packet is reverse-path-tested test all of the
default routes or just the route that is currently active given the load
balancing algorithm?

If only the one, current default route is used in the evaluation it
seems to me even in the most straightforward
dual-load-balanced-default-route configuration there is a race between
the time a packet is assigned an outgoing address & sent out the then
current default route and the routing code re-balancing and switching
the active default route (i.e prior to the reply packet -- or even in
the middle of active tcp connections).

The situation gets even worse (not even just a race condition) when you
apply policy routing to force the use of a particular default route.

Thots?

b.

> 
> Paul Viney
> 
> 
> On Wednesday 14 February 2007 21:17, Torsten Luettgert wrote:
> > This is one of my favourites :-)
> >
> > Usually that problem is caused by the rp_filter feature, which silently
> > drops packets that arrive on an interface answers wouldn't be routed to.
> >
> > Just try
> >
> > for i in /proc/sys/net/ipv4/conf/eth*/rp_filter; do
> > 	echo 0 > $i
> > done
> >
> > and see if that helps.
> > (indeed, you don't really need to switch it off for all of them, just
> > the uplink interfaces would be enough)
> >
> > Hth,
> > Torsten
> _______________________________________________
> LARTC mailing list
> LARTC@mailman.ds9a.nl
> http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
> 
> 
-- 
My other computer is your Microsoft Windows server.

Brian J. Murrell

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

[-- Attachment #2: Type: text/plain, Size: 143 bytes --]

_______________________________________________
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc

      parent reply	other threads:[~2007-02-18 20:30 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-02-13 13:50 [LARTC] Routing problem (RTNETLINK answers: Invalid argument) on Paul Viney
2007-02-13 20:40 ` Alex Samad
2007-02-13 21:54 ` Paul Viney
2007-02-14  2:53 ` Alex Samad
2007-02-14  7:30 ` Paul Viney
2007-02-14  7:35 ` Paul Viney
2007-02-14 20:17 ` [LARTC] Routing problem (RTNETLINK answers: Invalid argument) Torsten Luettgert
2007-02-14 23:30 ` [LARTC] Routing problem (RTNETLINK answers: Invalid argument) on Paul Viney
2007-02-15  0:00 ` Alex Samad
2007-02-18 20:30 ` Brian J. Murrell [this message]

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=1171830634.7053.58.camel@pc.ilinx \
    --to=brian@interlinx.bc.ca \
    --cc=lartc@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.