All of lore.kernel.org
 help / color / mirror / Atom feed
* [LARTC] Some questions about DS routing
@ 2004-03-06 16:35 Ricardo Pedro
  0 siblings, 0 replies; only message in thread
From: Ricardo Pedro @ 2004-03-06 16:35 UTC (permalink / raw)
  To: lartc

[-- Attachment #1: Type: text/plain, Size: 174 bytes --]

Hi, Im doing a research in IPv6 / DiffServ / Routing and I can't find 
any answers at others mailing lists.

Is anyone here that ant hint for me.

Best Regards
Ricardo Pedro

[-- Attachment #2: [Linux Diffserv] IPv6 - Ingress qdisc - Dsmark --]
[-- Type: message/rfc822, Size: 4053 bytes --]

From: Ricardo Pedro <ricardo@soferragens.pt>
To: kuznet@ms2.inr.ac.ru,  diffserv-general@lists.sourceforge.net
Subject: [Linux Diffserv] IPv6 - Ingress qdisc - Dsmark
Date: Sat, 06 Mar 2004 02:04:43 +0000
Message-ID: <404931BB.6030306@soferragens.pt>

Doing more tests I think that the problem of TC is with this combination.

For root ingress and in the downstream direction, everything works ok, 
but in the downstream no.

Once again, I'm asking for help, if the problem is mine, or this is a 
known bug!!!

Ricardo Pedro


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Diffserv-general mailing list
Diffserv-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/diffserv-general

[-- Attachment #3: [Linux Diffserv] dsfield in IPv6 --]
[-- Type: message/rfc822, Size: 4097 bytes --]

From: Ricardo Pedro <ricardo@soferragens.pt>
To: diffserv-general@lists.sourceforge.net
Subject: [Linux Diffserv] dsfield in IPv6
Date: Fri, 05 Mar 2004 21:24:38 +0000
Message-ID: <4048F016.4060706@soferragens.pt>

I was looking to the linux IPv6 code and I have another question:

-Why doesn't the ipv6 header code (ipv6.h) is correct to be conform to 
the IPv6 standard:

Traffic Class
Flow Label

I know that you can use get_dsfield() and change_dsfield() in ipv6, but 
the header structure still uses the 32 bits flowlabel.

Regards
Ricardo Pedro


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Diffserv-general mailing list
Diffserv-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/diffserv-general

[-- Attachment #4: [Linux Diffserv] Ingress qdisc in IPv6 --]
[-- Type: message/rfc822, Size: 5700 bytes --]

From: Ricardo Pedro <ricardo@soferragens.pt>
To: kuznet@ms2.inr.ac.ru,  diffserv-general@lists.sourceforge.net
Subject: [Linux Diffserv] Ingress qdisc in IPv6
Date: Fri, 05 Mar 2004 20:29:12 +0000
Message-ID: <4048E318.5040601@soferragens.pt>

Hi,

I'm trying to build a diffserv testbed and I want to mark the DSCP field 
in the IPv6 packet at the ingress router.

I'm using the folowing script

/sbin/tc qdisc add dev eth0 handle ffff:0 ingress
/sbin/tc filter add dev eth0 parent ffff:0 protocol ipv6 prio 0xffff u32 
ht 0x800:: order 0xfff match ip6 src ::/0  flowid 0x80
/sbin/tc qdisc add dev eth0 handle 1:0 root dsmark indices 256 
default_index 0x08
/sbin/tc filter add dev eth0 parent 1:0 protocol ipv6 prio 0xffff 
tcindex mask 0xff shift 0 fall_through
/sbin/tc class change dev eth0 classid 1:0x00 dsmark mask 0x3 value 0x68
/sbin/tc class change dev eth0 classid 1:0x04 dsmark mask 0x3 value 0xb8
/sbin/tc class change dev eth0 classid 1:0x08 dsmark mask 0x3 value 0x68
/sbin/tc class change dev eth0 classid 1:0x11 dsmark mask 0x3 value 0x28
/sbin/tc class change dev eth0 classid 1:0x12 dsmark mask 0x3 value 0x30
/sbin/tc class change dev eth0 classid 1:0x13 dsmark mask 0x3 value 0x38
/sbin/tc class change dev eth0 classid 1:0x21 dsmark mask 0x3 value 0x48
/sbin/tc class change dev eth0 classid 1:0x22 dsmark mask 0x3 value 0x50
/sbin/tc class change dev eth0 classid 1:0x23 dsmark mask 0x3 value 0x58
/sbin/tc class change dev eth0 classid 1:0x31 dsmark mask 0x3 value 0x68
/sbin/tc class change dev eth0 classid 1:0x32 dsmark mask 0x3 value 0x70
/sbin/tc class change dev eth0 classid 1:0x33 dsmark mask 0x3 value 0x78
/sbin/tc class change dev eth0 classid 1:0x41 dsmark mask 0x3 value 0x88
/sbin/tc class change dev eth0 classid 1:0x42 dsmark mask 0x3 value 0x90
/sbin/tc class change dev eth0 classid 1:0x43 dsmark mask 0x3 value 0x98


This script is not working for upstream traffic.
The strange is that this script works for downstream stream.

I think that the IPv6 code is not complete in the TC script, TC filters 
and other files.

If this is correct can anyone give me a hint how can a try to finish 
this code.

Best Regards
Ricardo Pedro


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Diffserv-general mailing list
Diffserv-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/diffserv-general

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2004-03-06 16:35 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2004-03-06 16:35 [LARTC] Some questions about DS routing Ricardo Pedro

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.