mpls-linux-general — Mailing list for general discussion about MPLS for Linux

[mpls-linux-general] mpls-linux 1.108 (BETA!!!)

[James R. L. <jl...@mi...>]

I just updated the CVS on sourceforge.  Check out the mpls-linux module
with a revision of mpls-linux_1_108.  This will have a kernel patch
a patch to iptables, and a patch to ppp.  You only need the iptables
patch if you want to use the new "instructions".

The basic functionality has been well tested, the new functionality
only has limited testing.

You'll notice this is the first release to have iptables.  Eventually
iptables will be the interface to bind a FEC to a LSP, not via the -B -f -O
combintion on mplsadm.

New instructions via mplsadm (only usable via -i and -o):

push:gen:16:eth1
fwd:gen:16:eth1
nffwd:0xF:0x1:0x44001001[:0x2:0x44001001]...
    where 0xF is the mask to apply to the nfmark or the packet, the
    resulting value is use for the lookup

    0x1:0x44001001 means packets with nfmark 0x1 (after mask is applied)
    is handled to the outgoing label entry with the key of 0x44001001
    (get the key from mpls_out)

    add upto 64 nfmark:key pairs

dsfwd:0xF:0x1:0x44001001[:0x2:0x44001001]...
    same as nfmark but the lookup value is taken from the ip header
    (this instruction is only valid on ingress LER)

expfwd:0x1:0x44001001[:0x2:0x44001001]...
    similar to nfmark but the lookup value is taken from the expbits
    in the incoming label shim.  There is not a mask value, only 8 exp:key
    pairs are allowed.

outgoing version 
set:eth1:ipv4:1.1.1.1
    set the outgoing next hop info

incoming version 
set:eth1
    set the rx-interface on the packet

set_ds:0x3
    set the dscp on the outgoing IP packet (only valid on egress LER)

set_exp:0x7
    set the expbits in the outgoing label shim

exp2tc:0x7:0x5[:0x4:0x100]...
    translate an expbit value from an incoming label shim into a
    tc-index which can be used by the tcindex traffic scheduler
    8 exp:tc pairs allowed

exp2ds:0x7:0x1:[0x1:0x6]...
    same at exp2tc except the dscp in the ip header is overwritten
    (only valid on egress LER)

tc2exp:0x5:0x7[:0x3F:0x4]...
    reverse of exp2tc, use a tc value on a packet and set the outgoing
    expbits in the label shim
    (64 tc:exp pairs allowed)

ds2exp:0x1:0x7[:0x6:0x1]...
    reverse of exp2ds, use a ds value in an IP header and set the
    outgoing expbits (only valid on ingress LER)
    (64 ds:exp pairs allowed)

NOTES:
    order does matter!
    don't think of outgoing info as having to push a label!
    I know some places I use key values and other places I use the
      old gen:16:eth1 values, evetually everything will use the keys

Please e-mail me with bugs and questions

Jim
-- 
James R. Leu

Re: [mpls-linux-general] Routing Table Changes

[Abhijit G. <gab...@ee...>]

Well,

If you want to check how routing updates are done to the kernel, look
through the sources of iproute2 package, which i think you must have done
already. The mechanism is based on netlink sockets. Look at

http://qos.ittc.ukans.edu/netlink/html/

for internals of Netlink sockets.

If you want to add some information to the kernel routing table. Please
ensure that whether it does not exist? As far as I know the kernel routing
implementation is quite exaustive. So a job of modifying it is certainly
non-trivial (and in most cases unnecessary.). There is a file in the
Documentation directory called networking/routing.txt . You can read that
to see what exactly is supported.

-abhijit

kanwardeep singh wrote :

>Hi All,
> We are doing a project which reqiures us to make a
>small network of PCs (linux 7.1), where each is
>behaving as a router.
> We then 've to implement MPLS network over this. What
>we want is..
> When network turns up all the router ard able to
>share their routing information (Dynamically)
> with each other. We've read that Linux uses "routed"
>deamon for this process, which in turn follows RIP,
>for
> dynamic routing. Till here its Ok... The problem is
>that we want to change the way routing
> informatio is stored in Linux kernel (i.e. we want to
>make some changes in routing tabel, by adding
> some more information to it, which is passed on by
>us)
> Can anyone tell us how is it possible, to modify
>existing routing table at runtime.
> Or can anyone tell which .c or .h files are
>responsible for calculation of routing table so that
> we can incorporate our changes in those files.
> Thanks in advance,
> Regards,
> Kanwar.
>
>
>
>__________________________________________________
>Do You Yahoo!?
>Send FREE video emails in Yahoo! Mail!
>http://promo.yahoo.com/videomail/
>
>_______________________________________________
>mpls-linux-general mailing list
>mpl...@li...
>https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
>

-- 
How do u delete a letter in vi and paste it after the next one?

"xp"

--

[mpls-linux-general] Routing Table Changes

[kanwardeep s. <kan...@ya...>]

Hi All,
 We are doing a project which reqiures us to make a
small network of PCs (linux 7.1), where each is
behaving as a router.
 We then 've to implement MPLS network over this. What
we want is..
 When network turns up all the router ard able to
share their routing information (Dynamically)
 with each other. We've read that Linux uses "routed"
deamon for this process, which in turn follows RIP,
for
 dynamic routing. Till here its Ok... The problem is
that we want to change the way routing
 informatio is stored in Linux kernel (i.e. we want to
make some changes in routing tabel, by adding
 some more information to it, which is passed on by
us)
 Can anyone tell us how is it possible, to modify
existing routing table at runtime.
 Or can anyone tell which .c or .h files are
responsible for calculation of routing table so that
 we can incorporate our changes in those files.
 Thanks in advance,
 Regards,
 Kanwar.



__________________________________________________
Do You Yahoo!?
Send FREE video emails in Yahoo! Mail!
http://promo.yahoo.com/videomail/

Re: [mpls-linux-general] writting into proc file

[Abhijit G. <gab...@ee...>]

Look for documentation of sysctl interface. It essentially does that..

But better use netlink if u want to pass data between userspace n
kernelspace. As far as i know sysctl interface has few limitations.

-abhijit

vvk...@hs... wrote :

>
>
>Hi  Folks !
>     Can anybody explain how will be the  mechanism to write into file
>   in the /proc filesystem from kernel space ...........
>
> In the 2.4.x linux kernel , for less than   PAGE_SIZE data transfer ,
>i am using the following mechanism to register the entry in the  /PROC
>filesytem...
>
>    struct proc_dir_entry *entry=create_proc_entry("xyz", S_IRUGO | S_IWUGO ,
>NULL );
>
>    if   (  entry )
>    {
>           entry->read_proc = xyz_read_proc;
>           entry->write_proc = xyz_write_proc;
>     }
>
>are these mean to say that accessing this file from userspace........or kernel
>space..........
>
>Is this mean to say that , when user invokes  "read" or "write" system directly
>or indirectly with ANSI library
>
>will be available for user to access this file ?
>
>I want to write the content into a file ..i.e in /proc  ....  from  kernel
>space.........
>
>What will be various options to acheive this ?
>
>can anybody send a small template for this , will be greatly
>appreciated..........
>
>Thanks in advance.........
>
>With  thanx and regards,
>Ramakrishna kvv.

Re: [mpls-linux-general] writting into proc file

[James R. L. <jl...@mi...>]

Take a look at how the entries in /proc/sys/ are handled.  I know the ipv4
stack has a couple of those.

Jim

On Fri, Jan 04, 2002 at 02:28:54PM +0530, vvk...@hs... wrote:
> 
> 
> Hi  Folks !
>      Can anybody explain how will be the  mechanism to write into file
>    in the /proc filesystem from kernel space ...........
> 
>  In the 2.4.x linux kernel , for less than   PAGE_SIZE data transfer ,
> i am using the following mechanism to register the entry in the  /PROC
> filesytem...
> 
>     struct proc_dir_entry *entry=create_proc_entry("xyz", S_IRUGO | S_IWUGO ,
> NULL );
> 
>     if   (  entry )
>     {
>            entry->read_proc = xyz_read_proc;
>            entry->write_proc = xyz_write_proc;
>      }
> 
> are these mean to say that accessing this file from userspace........or kernel
> space..........
> 
> Is this mean to say that , when user invokes  "read" or "write" system directly
> or indirectly with ANSI library
> 
> will be available for user to access this file ?
> 
> I want to write the content into a file ..i.e in /proc  ....  from  kernel
> space.........
> 
> What will be various options to acheive this ?
> 
> can anybody send a small template for this , will be greatly
> appreciated..........
> 
> Thanks in advance.........
> 
> With  thanx and regards,
> Ramakrishna kvv.
> 
> 
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu

[mpls-linux-general] writting into proc file

[<vvk...@hs...>]

Hi  Folks !
     Can anybody explain how will be the  mechanism to write into file
   in the /proc filesystem from kernel space ...........

 In the 2.4.x linux kernel , for less than   PAGE_SIZE data transfer ,
i am using the following mechanism to register the entry in the  /PROC
filesytem...

    struct proc_dir_entry *entry=create_proc_entry("xyz", S_IRUGO | S_IWUGO ,
NULL );

    if   (  entry )
    {
           entry->read_proc = xyz_read_proc;
           entry->write_proc = xyz_write_proc;
     }

are these mean to say that accessing this file from userspace........or kernel
space..........

Is this mean to say that , when user invokes  "read" or "write" system directly
or indirectly with ANSI library

will be available for user to access this file ?

I want to write the content into a file ..i.e in /proc  ....  from  kernel
space.........

What will be various options to acheive this ?

can anybody send a small template for this , will be greatly
appreciated..........

Thanks in advance.........

With  thanx and regards,
Ramakrishna kvv.

[mpls-linux-general] Query about rt_set_nexthop

[Jagdish <jag...@de...>]

Hi
I am studying the mpls 1.0 version and I have a query regarding the mpls
patch applied on rt_set_nexthop method(route.c).

1) The patch is applied to if (fi)  part of the method. i.e the mpls code is
executed only fib_result->fi structure is not null.
     The patch is not applied to else part.  How we are making sure that fi
should not be null in case of mpls forwarding.

2) In the if(fi) loop there is a condition check
  if (fi->fib_mtu == 0) {
  The code is assigning the pmtu to 576 or whatever relevant if this
condition true. Otherwise no assignment of the pmtu taken place.
   There is no else block corresponding to the if. The mpls patch is applied
after if loop and it decrements the mtu size.
   If our control comes from the if loop then it seems ok to me as we are
assigning pmtu in each invocation.
   But if fib_mtu !=0 then the patch is decrementing the rtable->u.dst.pmtu
in each invocation( I feel rtable is static ).
   Please help me.

Regards
jagdish

[mpls-linux-general] problem installing packages in ldp-portable-0.060

[anish v. <ani...@re...>]

Hi All,=0D=0A  We'r facing some problems in installing, ldp_linux=0D=0AWe '=
ve successfuly installed "mplsadm" utiliy in /usr/src/mpls-linux-1.0/util/=
=0D=0A=0D=0A1.=0D=0A  The problem is that in dir /usr/local/ldp-portable-0.=
0.60/ when I give=0D=0A# make=0D=0Afor some time make runs successfully... =
but after some time... error is generated :=0D=0Art ; make ldp_linux.o=0D=
=0Amake[1]: Entering directory `/usr/src/ldp-portable-0.060/port'=0D=0Amake=
[1]: `ldp_linux.o' is up to date.=0D=0Amake[1]: Leaving directory `/usr/src=
/ldp-portable-0.060/port'=0D=0Acd port ; make libldpport.a=0D=0Amake[1]: En=
tering directory `/usr/src/ldp-portable-0.060/port'=0D=0Amake[1]: `libldppo=
rt.a' is up to date.=0D=0Amake[1]: Leaving directory `/usr/src/ldp-portable=
-0.060/port'=0D=0Acd lib ; make libldp.a=0D=0Amake[1]: Entering directory `=
/usr/src/ldp-portable-0.060/lib'=0D=0Amake[1]: `libldp.a' is up to date.=0D=
=0Amake[1]: Leaving directory `/usr/src/ldp-portable-0.060/lib'=0D=0Agcc -o=
 ldp_linux port/ldp_linux.o lib/libldp.a port/libldpport.a lib/libldp.a=0D=
=0Acollect2: ld terminated with signal 11 [Segmentation fault]=0D=0Amake: *=
** [ldp_linux] Error 1=0D=0A-----------------------------------------------=
--=0D=0A2.=0D=0Aalso I downloaded all files of /linux-port dir from CVS & p=
ut them in /usr/src/ldp-portable-0.060/linux-port/=0D=0Awhen I give=0D=0A# =
make =0D=0Aor=0D=0A# make ldp_linux =0D=0Acc -I../lib -I. -Wall -g -O2 -D__=
LDP_LINUX__ -DMPLS_LINUX   -c -o=0D=0Aldp_linux.o ldp_linux.c n file includ=
ed from ../lib/ldp_struct.h:292,=0D=0A                 from ldp_trace.h:5,=
=0D=0A                 from ldp_linux.c:19:=0D=0A../lib/ldp_nortel.h:194:8:=
 warning: extra tokens at end of #endif=0D=0Adirective=0D=0Aldp_linux.c: In=
 function `show_route':=0D=0Aldp_linux.c:470: warning: implicit declaration=
 of function=0D=0A`ldp_tree_dump'=0D=0Aldp_linux.c: In function `main':=0D=
=0Aldp_linux.c:843: `LDP_ROUTE_UPDATE' undeclared (first use in this=0D=0Af=
unction)=0D=0Aldp_linux.c:843: (Each undeclared identifier is reported only=
 once=0D=0Aldp_linux.c:843: for each function it appears in.)=0D=0Amake: **=
* [ldp_linux.o] Error 1=0D=0A----------------------------------------=0D=0A=
 =0D=0AKindly Give proper guidence...in this regard,=0D=0AThanks in addvanc=
e, & a very Happy New Year To all=0D=0AAnish. =0A

Re: [mpls-linux-general] Problem Installing ldp_linux

[James R. L. <jl...@mi...>]

What errors?  Most likely it is complaining about not finding the mpls.h
header file.  Take a look at the mpls-linux CVS respository on sf.
There should be a file "README.1st.really" (it was added after releasing
mpls-linux 1.0) that will help you fix this.

Jim


On Mon, Dec 31, 2001 at 02:58:48AM -0800, kanwardeep singh wrote:
> Hi All,
>   We've compiled linux-2.4.13 with after applying
> patch linux-mpls-1.0/linux-mpls.diff from project
> Linux MPLS at sourcefourge. 
>   We've compiled mplasdm successfully.
> Now when I try to "make" contents in directory
> linux-port (taken from CVS)its giving errors....(to
> impliment the test mentioned in README.sample) 
> I've placed linux-port directory in
> /root/sf/ldp-portable-0.0.60/
> Please, give solution to this problem,
> Regards & Happy new year,
> Kanwar.
> 
> __________________________________________________
> Do You Yahoo!?
> Send your FREE holiday greetings online!
> http://greetings.yahoo.com
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu

Re: [mpls-linux-general] How to update a common ATM switch to a ATM-LSR?

[James R. L. <jl...@mi...>]

On Fri, Dec 28, 2001 at 10:23:53AM +0800, =D5=C5=CE=B0=D2=BB wrote:
> Hi,
>   I am a novice at MPLS,I'd appreciate any help to my question.
> I want to program to update a common ATM switch to a ATM-LSR,I think th=
e
> primary issue is to put the allocated label into the ATM cell header as
> the VPI/VCI.But I do not know whether my idea is available and how to
> implement it.
> I really need your help!!!

I do not understand what you are trying to do.  Try to explain it in a li=
ttle
more detail and I'll see if I can help.

Jim

>=20
> Any help would be greatly appreciated.
>=20
> Regards.
>=20
> Weiyi Zhang
>=20
>=20
>=20
>=20
>=20
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> http://vip.163.com          =D7=A2=B2=E1VIP=D3=CA=CF=E4  =B4=F3=BD=B1=C8=
=CE=C4=E3=C4=C3
> http://im.163.com           =CD=F8=D2=D7=B6=CC=D0=C5=CD=A8=A3=BA=C3=E2=B7=
=D1=B7=A2=CB=CD=CA=D6=BB=FA=B6=CC=D0=C5=B5=C4=D7=C0=C3=E6=C8=ED=BC=FE
> http://dating.163.com       =D4=BC=BB=E1=D6=AE=B3=C7=A3=AC=B4=B9=C1=AF=C4=
=E3=D1=DB=D6=D0=CE=DE=C9=F9=C4=C5=BA=B0=A3=A1
>=20
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

--=20
James R. Leu

Re: [mpls-linux-general] need a tool to test my mpls application

[James R. L. <jl...@mi...>]

Not that I know of.  You could possibly modify pktgen to send MPLS frames.

Jim

On Wed, Dec 26, 2001 at 12:46:15PM +0530, Jagdish wrote:
> Hi,
> I have written an application to receive mpls packets. Now I want to test my
> application. Is there any network simulator/application / tool available
> which can send/receive mpls packets.
> 
> regards
> Jagdish
> 
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu

Re: [mpls-linux-general] MPLS+IPSec

[James R. L. <jl...@mi...>]

Thanks for the words of encoragment.

On Sat, Dec 22, 2001 at 02:08:44PM +0100, Vito Covito wrote:
> We are Naple's University Federico II students.
> First we want congratulate for the mpls project that you have developed.
> We are using the latest version (mpls-1.0) and we want use it with IPSEC to 
> create secure  VPNs. 
> Have you some advice for the question??

I have exchanged e-mail with the IPSEC people, but I do not have any
experience using IPSEC.

> You know if there is some on-line-documents for this purpose?

If you come up with a working implementation please post you results here.

Thanks,
Jim
-- 
James R. Leu

[mpls-linux-general] Problem Installing ldp_linux

[kanwardeep s. <kan...@ya...>]

Hi All,
  We've compiled linux-2.4.13 with after applying
patch linux-mpls-1.0/linux-mpls.diff from project
Linux MPLS at sourcefourge. 
  We've compiled mplasdm successfully.
Now when I try to "make" contents in directory
linux-port (taken from CVS)its giving errors....(to
impliment the test mentioned in README.sample) 
I've placed linux-port directory in
/root/sf/ldp-portable-0.0.60/
Please, give solution to this problem,
Regards & Happy new year,
Kanwar.

__________________________________________________
Do You Yahoo!?
Send your FREE holiday greetings online!
http://greetings.yahoo.com

[mpls-linux-general] =?gb2312?B?SG93IHRvIHVwZGF0ZSBhIGNvbW1vbiBBVE0gc3dpdGNoIHRvIGEgQVRNLUxTUj8=?=

[ÕÅΰһ <zh...@16...>]

Hi,
  I am a novice at MPLS,I'd appreciate any help to my question.
I want to program to update a common ATM switch to a ATM-LSR,I think the
primary issue is to put the allocated label into the ATM cell header as
the VPI/VCI.But I do not know whether my idea is available and how to
implement it.
I really need your help!!!

Any help would be greatly appreciated.

Regards.

Weiyi Zhang





=============================================================
http://vip.163.com          ×¢²áVIPÓÊÏä  ´ó½±ÈÎÄãÄÃ
http://im.163.com           ÍøÒ׶ÌÐÅͨ£ºÃâ·Ñ·¢ËÍÊÖ»ú¶ÌÐŵÄ×ÀÃæÈí¼þ
http://dating.163.com       Ô¼»áÖ®³Ç£¬´¹Á¯ÄãÑÛÖÐÎÞÉùÄź°£¡

Re: [mpls-linux-general] need a tool to test my mpls application

[Saeed A. <she...@ho...>]

Hello Jagdish,

You might need to use any of the following;

- Ixia
- SmartBit
- Router Tester

These are traffic generators for MPLS/RSVP/LDP and Routing Protocols too.

Hope this will help,

Saeed


_________________________________________________________________
Join the world’s largest e-mail service with MSN Hotmail. 
http://www.hotmail.com

[mpls-linux-general] need a tool to test my mpls application

[Jagdish <jag...@de...>]

Hi,
I have written an application to receive mpls packets. Now I want to test my
application. Is there any network simulator/application / tool available
which can send/receive mpls packets.

regards
Jagdish

[mpls-linux-general] MPLS+IPSec

[Vito C. <vi...@pt...>]

We are Naple's University Federico II students.
First we want congratulate for the mpls project that you have developed.
We are using the latest version (mpls-1.0) and we want use it with IPSEC to 
create secure  VPNs. 
Have you some advice for the question??
You know if there is some on-line-documents for this purpose?
Tanks for all.
and Merry Christmas
Buon Natale
Vito e Carlo

ps. Excuse for my crippled english

Re: [mpls-linux-general] how Ip will give the packet to MLS layer?

[James R. L. <jl...@mi...>]

README.ingress explains how to do ingress policing.

Look at README.example in the mplsadm/utils directory

On Fri, Dec 21, 2001 at 05:10:35AM -0000, ramakrishna kvv wrote:
> 
> Hello  All   
> 	We are using mpls-linux code with linux 2.4.1 kernel. we configured the mpls domain and trying to send 
> a packet from host , conmnected to LER.
> 
>   In the  "README.ingress " file , I think James wrote that, specified how to configure the LER.
> 
>    We are trying to simulate the example, that james explained,  in our college lab.
> 
>        We want to know how the packet will be delivered to MPLS at LER's ?,  Since at LER unlabeled packet will come  and ethernet driver gives it to IP. Then how IP will decide to give the packet to MPLS layer?
> 
>     Can anybody explain or give any pointer to that. 
> 
>   How the IP will give the packet to mpls layer , whether it wants to send as labeld packet ?
> 
>  Thanks...
>  Regards,  Ramki. 
> 
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu

Re: [mpls-linux-general] backports??

[James R. L. <jl...@mi...>]

On Wed, Dec 19, 2001 at 07:44:46PM +0000, Chip Rodden wrote:
> 
> 
> Hi Jim,
> 
> I haven't been keeping up with the MPLS list for a while now so
> I'm not sure if this as been discussed.  In any case, are you or
> is anyone else aware of a backport of a fairly recent release of your
> stuff to a 2.2 kernel?

Not that I'm aware of.  If you attemtp this yourself, I'd be more then
happy to assist with any problems you run into.

Jim

> 
> Thanks,
> 
> Chip
> 
> _________________________________________________________________
> MSN Photos is the easiest way to share and print your photos: 
> http://photos.msn.com/support/worldwide.aspx
> 
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu
jl...@mi...

Re: [mpls-linux-general] how Ip will give the packet to MLS layer?

[Amit K. <ami...@ya...>]

please check the post few days ago for  the info 
in the thread 
"Looking for examples to test TC/DS/IPTABLE"


-amit
--- ramakrishna kvv <ram...@re...>
wrote:
> 
> Hello  All   
> 	We are using mpls-linux code with linux 2.4.1
> kernel. we configured the mpls domain and trying to
> send 
> a packet from host , conmnected to LER.
> 
>   In the  "README.ingress " file , I think James
> wrote that, specified how to configure the LER.
> 
>    We are trying to simulate the example, that james
> explained,  in our college lab.
> 
>        We want to know how the packet will be
> delivered to MPLS at LER's ?,  Since at LER
> unlabeled packet will come  and ethernet driver
> gives it to IP. Then how IP will decide to give the
> packet to MPLS layer?
> 
>     Can anybody explain or give any pointer to that.
> 
> 
>   How the IP will give the packet to mpls layer ,
> whether it wants to send as labeld packet ?
> 
>  Thanks...
>  Regards,  Ramki. 
> 
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
>
https://lists.sourceforge.net/lists/listinfo/mpls-linux-general


__________________________________________________
Do You Yahoo!?
Check out Yahoo! Shopping and Yahoo! Auctions for all of
your unique holiday gifts! Buy at http://shopping.yahoo.com
or bid at http://auctions.yahoo.com

[mpls-linux-general] how Ip will give the packet to MLS layer?

[ramakrishna k. <ram...@re...>]

Hello  All   =0D=0A	We are using mpls-linux code with linux 2.4.1 kernel. w=
e configured the mpls domain and trying to send =0D=0Aa packet from host , =
conmnected to LER.=0D=0A=0D=0A  In the  "README.ingress " file , I think Ja=
mes wrote that, specified how to configure the LER.=0D=0A=0D=0A   We are tr=
ying to simulate the example, that james explained,  in our college lab.=0D=
=0A=0D=0A       We want to know how the packet will be delivered to MPLS at=
 LER's ?,  Since at LER unlabeled packet will come  and ethernet driver giv=
es it to IP. Then how IP will decide to give the packet to MPLS layer?=0D=
=0A=0D=0A    Can anybody explain or give any pointer to that. =0D=0A=0D=0A =
 How the IP will give the packet to mpls layer , whether it wants to send a=
s labeld packet ?=0D=0A=0D=0A Thanks...=0D=0A Regards,  Ramki. =0A

[mpls-linux-general] how Ip will give the packet to MLS layer?

[ramakrishna k. <ram...@re...>]

Hello  All   =0D=0A	We are using mpls-linux code with linux 2.4.1 kernel. w=
e configured the mpls domain and trying to send =0D=0Aa packet from host , =
conmnected to LER.=0D=0A=0D=0A  In the  "README.ingress " file , I think Ja=
mes wrote that, specified how to configure the LER.=0D=0A=0D=0A   We are tr=
ying to simulate the example, that james explained,  in our college lab.=0D=
=0A=0D=0A       We want to know how the packet will be delivered to MPLS at=
 LER's ?,  Since at LER unlabeled packet will come  and ethernet driver giv=
es it to IP. Then how IP will decide to give the packet to MPLS layer?=0D=
=0A=0D=0A    Can anybody explain or give any pointer to that. =0D=0A=0D=0A =
 How the IP will give the packet to mpls layer , whether it wants to send a=
s labeld packet ?=0D=0A=0D=0A Thanks...=0D=0A Regards,  Ramki. =0A

[mpls-linux-general] backports??

[Chip R. <chi...@ho...>]

Hi Jim,

I haven't been keeping up with the MPLS list for a while now so
I'm not sure if this as been discussed.  In any case, are you or
is anyone else aware of a backport of a fairly recent release of your
stuff to a 2.2 kernel?


Thanks,

Chip

_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx

[mpls-linux-general] LDP-Portable with ATM switch - Query abt code.

[<nc...@si...>]

Hi Mr Leu,

  I have tried out the latest ldp-portable with an MPLS enabled ATM switch. I performed the following steps:

  1) add global "src address".
  2) add peer "switch's IP address".

  When I sniffed on the wire using Ethereal, I noticed that  the ATM switch upon receiving the Linux router's "Hello" message, initiated a TCP connection and sent a "Initialization" message to it with ATM session parameters.

 However, ldp-portable failed to respond. I delved into the code and found out that in the function ldp_stat_recv_init (from ldp_state_funcs), the function ldp_global_find_adj_ldpid returned a NULL value.

  Looking further I found that ldp_global_find_adj_ldpid called LDP_LIST_HEAD macro which produces a NULL value.

   I do not really understand why passing &g->adj into LDP_LIST_HEAD, where I think g turns out to be a global variable "cfg_handle", will produce a NULL value...

   I will appreciate it if u can advise me on this issue.. Anyway I can get around it?

   Thanks a lot!

Regards,

Paul

 


>>Date:    Tue, 18 Dec 2001 08:54:28 -0600
>>From:    James R. Leu <jl...@mi...>
>>To:      ncc49a <nc...@si...>
>>CC:      mpl...@li...; eng...@nu...
>>Subject: Re: [mpls-linux-general] LDP (linux-port)
>>
Sorry for not getting back to you earlier.  Did someone help you
out with this?  Mostly likely you don't have MPLS_LINUX defined in the
makefile.

Jim

On Tue, Dec 11, 2001 at 05:52:05PM +0800, ncc49a wrote:
> Hi everybody,
> 
>    I have downloaded the latest ldp implementation from the 
> CVS. I have followed the instructions exactly as specified 
> in the README in /linux-port for a direct connection 
> between 2 PCs.
> 
>    When done, a ldp session seems to have been established 
> between the 2 linux LSRs. Showing "database", "session" 
> and "route" all yield some results.
> 
>    However, when I tried to 
> ead /proc/mpls_in, /proc/mpls_out, /proc/mpls_fec files, 
> they were all blank, contrary to the README.
> 
>    That means no label bindings have been established 
> between the 2 LSRs right? 
> 
>    From the mailing list, I have noticed Mr Kauder Thorsten 
> is doing something similiar at the moment. I wonder if you 
> faced such an issue.
> 
>    I guess I must have missed out something. I will 
> appreciate all help/advice provided. 
> 
>    Thanks a lot!
> 
> Regards,
> 
> Paul
> 
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu
jl...@mi...


_________________________________________________
The simple way to read all your emails at ThatWeb
http://www.thatweb.com

Re: [mpls-linux-general] Looking for examples to test TC/DS/IPTABLE

[Olivier D. <Oli...@rd...>]

Hi Jim,

Comments below ...

James R. Leu wrote:

> Comments below ....
> 
> On Tue, Dec 18, 2001 at 04:21:48PM +0100, Olivier Dugeon wrote:
> 
>>Hi Jim,
>>
>>James R. Leu wrote:
>>
>>
>>>I see the light now.....
>>>
>>>Last night I dug around the kernel and I now see why you think that
>>>netfilter is the best way for interacting with the ipv4 routing table.
>>>I will look more at your work and see how I can make it less MPLS specific.
>>>
>>>Right now I think I am going to add a new netlink.  POST_ROUTING_SLOW.
>>>This will allow a netlink to modify the route cache entry.  The result will
>>>be that the netlink code for this will only be run for the first packet
>>>in the "flow".    The rest will hit the entry in the route cache and
>>>will be redirected to the MPLS layer.
>>>
>>>
>>
>>Ok, i think it's a good way setup the process of the first packet. It's 
>>that we do by hacking the rt_set_nexthop code. This function is call 
>>only for the first packet of a flow (ie. the first ping packet). Then, 
>>the entry are kept into the route cash a moment (so, stoping a ping and 
>>start some time after a second ping, doesn't call another time 
>>rt_set_nexthop because the route cash entry is still valid).
>>
>>What do you intend to do with this new POST_ROUTING_SLOW netlink ?
>>
> 
> -change the pmtu
> -use the some index on the skb to lookup the moi, attach it to the route cache
>  entry
> -change rt->dst.output to be mpls_output
> 


Ok. fine.

> This will require that everytime a moi is deleted, the route cache must
> be cleared. :-(
> 


In our original code, the moi is not delete. The moi was setup with 
mplsadm but no route entry was added (ie. no fec was given to mplsadm 
and dst_aux_proto was not set)

The packet go to ip_route_input function first. Then, in the case of the 
first packet, fast route is not find (ie. there is no route cash entry 
corresponding to this flow). So, ip_route_input call ip_route_input_slow 
function which call rt_set_next_hop where reside the mpls processing 
stuff. After rt_set_nexthop, ip_route_input_slow finish to compute a new 
route cash entry and store it into the route cash table. So, the 
subsequent packet are only process by ip_route_input because a valid 
route cash entry is find. So, i think route cash entry is not cleared, 
only a new one is created.


> What do you think of this?
> 
> What should be done with skb that come through POST_ROUTING_SLOW but
> a moi could not be found?
> 
> 1. I could drop it, then a route cache entry will not be created and
> next time a packet for this flow is seen it will be slow routed again
> giving another chance for a valid moi to be found.
> 
> 2. I could let it pass and let it go via IP.  Also subsequent packets for
> this flow will be fast routed, thus not given a chance to find valid moi.
> To get around this everytime a new moi is created the route cache could
> be cleared, thuis forcing packets to go through the slow path again.
> 
> What do you suggest?
> 


2. If moi could not be found for the first packet, subsequent packet 
failed to. So, the better is to process this packet as usual ie. normal 
ip stuff routing without MPLS processing. But, don't drop it and don't 
clear the route cash entry. I think that if the first packet doesn't 
activate the MPLS processing, is that it don't do. If it failed because 
value are not correct (mplsadm, iptable ... miss configuration) it's an 
operator problem. So, saying that with an MPLS_DEBUG output is 
sufficient. I would add to mplsadm the iptable and tc configuration to 
avoid a such situation but haven't find any time to do it.

I think, POST_ROUTING_SLOW must be setup if you would use another value 
  (ie. nfmark or mpls_index) to retrieve the moi from this skb entry. I 
don't if need to be dynamic like you suggest or just a compile flag for 
the kernel like other CONFIG_xxx.

Your previous post about MPLS_OP_NF_FWD seems to be better. The kernel 
code (inside rt_set_nexthop) can dig with an mpls mark comming from the 
skb to retrieve the moi information. This mpls mark can be setup by what 
you want (iptable, ipchains, tc ...). The purpose of MPLS_OP_NF_FWD is 
just for mplsadm to setup a moi entry into the RADIX_TREE which 
correspond to this mpls mark (label value is user friendly). So, the 
normal processing become (pseudo code) :

if (mpls_mark)
     retrieve_moi_from_mpls_mark(mpls_mark);
else
     moi = proto_data[AUX_PROTO_DATA_MPLS];

Now, a better one is to merge the two approach. Only use mpls_mark. This 
mpls_mark can be setup by iptable, tc or other AND can be setup by the 
normal mpls stuff based only on the ip dst address. So, the process become:

retrieve_moi(mpls_mark ? mpls_mark : proto_data[AUX_PROTO_DATA_MPLS]);

proto_data doesn't contain a pointer to the moi, but the corresponding 
entry (ie. the label value)

What do think of this ?

Olivier
-- 
  FTR&D/DAC/CPN
  Technopole Anticipa     | mailto:Oli...@fr...
  2, Avenue Pierre Marzin | Phone:  +(33) 2 96 05 28 80
  F-22307 LANNION         | Fax:    +(33) 2 96 05 18 52