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

[mpls-linux-general] thanks for mpls

[cyb3rbyt3 <111...@st...>]

allow all thanks for all, discussion, open mind and all, now i finished 
my last final project in my university. by the way sorry for james r 
leu, if i change a lot of your script, it's because i need for my last 
final project
btw
thanks for all

[mpls-linux-general] mpls-bgp and mpls cr-ldp

[Dorin P. <dor...@gm...>]

Hello!

I could find informations about mpls-bgp (rfc3107) and mpls cr-ldp (3212)
implemantion stage. Can you help me in finding beta versions, partial
implementations or more informations about these topics?

--
Dorin Pena
CCNA, CCAI

Re: Re: [mpls-linux-general] command help

[Josdeyvi R. <jr...@uo...>]

Hi Kishor,

For the label table the MIB MPLS-LSR have 3 major tables: =
mplsInSegmentTable, mplsOutSegmentTable, mplsXCTable. To populate these =
tables, your SNMP Agent must read the ilm, nhlfe and xc internal tables =
from MPLS for Linux and put the things in the correct place.

For example, if I do a set of commands like:


# mpls labelspace add dev eth1 labelspace 0
# mpls ilm add label gen 100 labelspace 0

I will get the following ILM table:

[root@MPLSCORE~]# mpls ilm sh

ILM entry label gen 100 labelspace 0 proto ipv4=20

pop forward key 0x00000002 (385664 bytes, 6026 pkts, 0 dropped)

And my Labelspace will be:

[root@MPLSCORE~]# mpls labelspace sh

LABELSPACE entry dev eth1 labelspace 0

The SNMP agent then must READ this output in somehow. You can directly =
acess the kernel using RTNetlink protocol (hack the mpls command of MPLS =
for Linux to learn how to do it) ou simple read and scan the output =
lines in your program. Doing it you must have the following at your MIB =
browser:

[root@MPLSCORE~]# snmpwalk -v2c -c public -O T -O S -O n localhost =
MPLSLSRSTDMIB::mplsInSegmentTable

.1.3.6.1.2.1.10.166.2.1.4.1.2.4.0.0.0.1 =3D INTEGER: 0

.1.3.6.1.2.1.10.166.2.1.4.1.3.4.0.0.0.1 =3D Gauge32: 100

.1.3.6.1.2.1.10.166.2.1.4.1.4.4.0.0.0.1 =3D OID: .0.0

.1.3.6.1.2.1.10.166.2.1.4.1.5.4.0.0.0.1 =3D INTEGER: 1

.1.3.6.1.2.1.10.166.2.1.4.1.6.4.0.0.0.1 =3D INTEGER: ipV4(1)

.1.3.6.1.2.1.10.166.2.1.4.1.7.4.0.0.0.1 =3D HexSTRING:

34 2E 30 2E 30 2E 30 2E 32 00 FF BF AA D8 EB 43 [4.0.0.0.2......C]=20

.1.3.6.1.2.1.10.166.2.1.4.1.8.4.0.0.0.1 =3D INTEGER: other(2)

.1.3.6.1.2.1.10.166.2.1.4.1.9.4.0.0.0.1 =3D OID: .0.0

.1.3.6.1.2.1.10.166.2.1.4.1.10.4.0.0.0.1 =3D INTEGER: active(1)

.1.3.6.1.2.1.10.166.2.1.4.1.11.4.0.0.0.1 =3D INTEGER: volatile(2)



I recommend you to read the following book:



Nadeu, Thomas D., "MPLS Network Management", 2003, Morgan Kaufmann =
Publishers.



Good luck!



Josdeyvi Russi



----------------------------------------------
  The scope stuff is not mpls-linux specific, read the iproute2 =
documents
  for info about 'scope'.

  There is a SF project that implements a SNMP agent for mpls-linux.  Do
  a search on the SF site for mpls and it will be one of the matches.

  BTW SF =3D3D=3D3D sourceforge.net

  On Wed, Feb 22, 2006 at 10:24:27AM +0530, kis...@wi... =
wrote:
  > Hii
  > =3D20
  > =3D20
  > I didn't get what do mean scope link in
  > "
  >    10.200.19.1 via 10.200.19.1 dev eth0 spec_nh 0x8847 0x2
  >    192.168.2.9/24 dev eth0 scope link
  >    10.200.19.1/24 dev eth0 proto kernel scope link src =
10.200.19,.1=3D20
  >    127.0.0.0/8 dev lo scope link"
  > =3D20
  > One more thing how can I access label table from SNMP agent? =
Specifically=3D
   are there any functions or commands for this?. Help me
  > =3D20
  > Thanks and Regards,
  > =3D20
  > Kishor
  > =3D20

  --=3D20
  James R. Leu
  jl...@mi...

Re: [mpls-linux-general] problem running mpls

[Dorin P. <dor...@gm...>]

I've solved the problem!

Thank you for your support!



On 2/23/06, James R. Leu <jl...@mi...> wrote:
>
> On Thu, Feb 23, 2006 at 08:33:32AM +0200, Dorin Pena wrote:
> > Thank you for your response!
> >
> > Yes, the module was loaded before I gave the command.
> > As I said in my previous email, some mpls commands work fine:
> >
> > dorian:~# mpls nhlfe  show
> > NHLFE entry key 0x00000003 mtu 0 propagate_ttl
> >          (0 bytes, 0 pkts)
> >
> > dorian:~# mpls nhlfe change key 0x2 instructions push gen 10000 nexthop
> eth0
> > ipv4 141.85.99.1
>
> Change 'change' to 'add' and it will work fine.
>
> PS Please make sure all mpls-linux questions and replies go to the
> mpls-linux-general mailing list.
>
> > RTNETLINK answers: No such process
> >
> > I still get this error.
> > DO you have further advice?
> >
> > Thank you for your support!
> >
> > Best regards!
> >
> > On 2/22/06, James R. Leu <jl...@mi...> wrote:
> > >
> > > Are you using the kernel RPM I created?
> > > If so did you read the release notes?
> > > From the release notes:
> > >
> > > ******* NOTE **********
> > > If you are using one of my provided kernel RPMs, you must issue a
> > > 'modprobe mpls4' before attempting any 'mpls' commands.  This is
> because
> > > MPLS is compiled as a module and must be loaded!!!!
> > > ******* NOTE **********
> > >
> > > On Wed, Feb 22, 2006 at 01:21:58PM +0200, Dorin Pena wrote:
> > > > Hello!
> > > >
> > > > I have a problem with mpls kernel 2.6.15. I have install from
> sources
> > > > downloaded from SourceForge and I also installed iproute-2.6.11.
> When I
> > > > tried to configure o simple LSP (as the example shows), I got the
> > > following
> > > > errors:
> > > >
> > > > dorian:/usr/src# mpls ilm add label gen 100 labelspace 0
> > > > RTNETLINK answers: Cannot allocate memory
> > > >
> > > > dorian:/usr/src# mpls nhlfe change key 0x2 instructions push gen
> 10000
> > > > nexthop eth0 ipv4 141.85.99.1
> > > > RTNETLINK answers: No such process
> > > >
> > > > mpls show works fine..
> > > >
> > > > Can anyone help me, please?
> > > >
> > > >
> > > > --
> > > > Dorin Pena
> > > > CCNA, CCAI
> > >
> > > --
> > > James R. Leu
> > > jl...@mi...
> > >
> > >
> > >
> >
> >
> > --
> > Dorin Pena
> > CCNA, CCAI
>
> --
> James R. Leu
> jl...@mi...
>
>
>


--
Dorin Pena
CCNA, CCAI

Re: [mpls-linux-general] problem running mpls

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

Are you using the kernel RPM I created?
If so did you read the release notes?
=46rom the release notes:

******* NOTE **********
If you are using one of my provided kernel RPMs, you must issue a
'modprobe mpls4' before attempting any 'mpls' commands.  This is because
MPLS is compiled as a module and must be loaded!!!!
******* NOTE **********

On Wed, Feb 22, 2006 at 01:21:58PM +0200, Dorin Pena wrote:
> Hello!
>=20
> I have a problem with mpls kernel 2.6.15. I have install from sources
> downloaded from SourceForge and I also installed iproute-2.6.11. When I
> tried to configure o simple LSP (as the example shows), I got the followi=
ng
> errors:
>=20
> dorian:/usr/src# mpls ilm add label gen 100 labelspace 0
> RTNETLINK answers: Cannot allocate memory
>=20
> dorian:/usr/src# mpls nhlfe change key 0x2 instructions push gen 10000
> nexthop eth0 ipv4 141.85.99.1
> RTNETLINK answers: No such process
>=20
> mpls show works fine..
>=20
> Can anyone help me, please?
>=20
>=20
> --
> Dorin Pena
> CCNA, CCAI

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] problem running mpls

[Dorin P. <dor...@gm...>]

Hello!

I have a problem with mpls kernel 2.6.15. I have install from sources
downloaded from SourceForge and I also installed iproute-2.6.11. When I
tried to configure o simple LSP (as the example shows), I got the following
errors:

dorian:/usr/src# mpls ilm add label gen 100 labelspace 0
RTNETLINK answers: Cannot allocate memory

dorian:/usr/src# mpls nhlfe change key 0x2 instructions push gen 10000
nexthop eth0 ipv4 141.85.99.1
RTNETLINK answers: No such process

mpls show works fine..

Can anyone help me, please?


--
Dorin Pena
CCNA, CCAI

Re: [mpls-linux-general] command help

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

The scope stuff is not mpls-linux specific, read the iproute2 documents
for info about 'scope'.

There is a SF project that implements a SNMP agent for mpls-linux.  Do
a search on the SF site for mpls and it will be one of the matches.

BTW SF =3D=3D sourceforge.net

On Wed, Feb 22, 2006 at 10:24:27AM +0530, kis...@wi... wrote:
> Hii
> =20
> =20
> I didn't get what do mean scope link in
> "
>    10.200.19.1 via 10.200.19.1 dev eth0 spec_nh 0x8847 0x2
>    192.168.2.9/24 dev eth0 scope link
>    10.200.19.1/24 dev eth0 proto kernel scope link src 10.200.19,.1=20
>    127.0.0.0/8 dev lo scope link"
> =20
> One more thing how can I access label table from SNMP agent? Specifically=
 are there any functions or commands for this?. Help me
> =20
> Thanks and Regards,
> =20
> Kishor
> =20

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] command help

[<kis...@wi...>]

Hii
=20
=20
I didn't get what do mean scope link in
"
   10.200.19.1 via 10.200.19.1 dev eth0 spec_nh 0x8847 0x2
   192.168.2.9/24 dev eth0 scope link
   10.200.19.1/24 dev eth0 proto kernel scope link src 10.200.19,.1=20
   127.0.0.0/8 dev lo scope link"
=20
One more thing how can I access label table from SNMP agent? =
Specifically are there any functions or commands for this?. Help me
=20
Thanks and Regards,
=20
Kishor
=20

Re: [mpls-linux-general] command help

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

On Tue, Feb 21, 2006 at 12:45:33PM +0530, kis...@wi... wrote:
> Hi,
>          In command "ip route add 10.200.19.1/32 via 10.200.19.1 spec_nh =
0x8847 0x2" what do we mean spec_nh=20

spec_nh stands for special nexthop.  spec_nh was my first attempt at buildi=
ng
infrastructure for Ipv4|6 to interact with nexthops that had a different
protocol (IPSEC and MPLS are the two I had in mind).  In mpls-linux 1.950
spec_nh has gone away and the 'shim' system has been implemented.

spec_nh 0x8847 0x2 means special nexthop of type MPLS (0x8847 is the ethern=
et
protocol ID for MPLS unicast) and a MPLS NHLFE key of 0x2.

> When I give command "mpls nhlfe show? i get the following:
>  NHLFE entry key 0x00000002 mtu 1496 propagate_ttl
>    push gen 11111 set eth0 ipv4 10.200.19.1 (0 bytes, 0 pkts, 0 dropped)
> What do mean mtu 1496 propagate_ttl=20

MTU is the maximum packet size that can be sent on the NHLFE.  propagate_ttl
means that we copy the TTL from the IPv4|6 header if this node is acting
as a ingress LER.

> The "ip rote show "gives the following:
>    10.200.19.1 via 10.200.19.1 dev eth0 spec_nh 0x8847 0x2
>    192.168.2.9/24 dev eth0 scope link
> 10.200.19.1/24 dev eth0 proto kernel scope link src 10.200.19,.1=20
> 127.0.0.0/8 dev lo scope link
> what do mean scope link?
>=20
> Also what are the commands to create table? and access the table?
> Sorry for disturbing by these questions. As Iam new to these please help =
me.

Your last two questions are not specific to mpls-linux and can easily be
found via google or the like. (hint "lartc")

>=20
> Regards,
> Kishor=20
>=20
>=20
>                          =20

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] command help

[<kis...@wi...>]

Hi,
         In command "ip route add 10.200.19.1/32 via 10.200.19.1 spec_nh =
0x8847 0x2" what do we mean spec_nh=20

When I give command "mpls nhlfe show? i get the following:
 NHLFE entry key 0x00000002 mtu 1496 propagate_ttl
   push gen 11111 set eth0 ipv4 10.200.19.1 (0 bytes, 0 pkts, 0 dropped)
What do mean mtu 1496 propagate_ttl=20

The "ip rote show "gives the following:
   10.200.19.1 via 10.200.19.1 dev eth0 spec_nh 0x8847 0x2
   192.168.2.9/24 dev eth0 scope link
10.200.19.1/24 dev eth0 proto kernel scope link src 10.200.19,.1=20
127.0.0.0/8 dev lo scope link
what do mean scope link?

Also what are the commands to create table? and access the table?
Sorry for disturbing by these questions. As Iam new to these please help =
me.

Regards,
Kishor=20


                         =20

Re: [mpls-linux-general] Traffic Splitting based on destination ports

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

This question has been asked on the mailing list already, please check the
archives for details.  The general idea is that you have to use a modified
iptables with the spech_nh target. (in 1.950 it is called the mpls target).

On Mon, Feb 20, 2006 at 11:26:51AM -0500, Khac Nguyen wrote:
> Hi James:
>=20
> I am using mpls-linux version 1.946, and I am trying to split incoming=20
> traffic into different LSP paths. Can you help me with the instructions=
=20
> how to configure my ingress LER for this purpose? Thanks in advance.
>=20
> Regards,
>=20
> Khac Nguyen
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=
=3D121642
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] Traffic Splitting based on destination ports

[Khac N. <ktn...@co...>]

Hi James:

I am using mpls-linux version 1.946, and I am trying to split incoming 
traffic into different LSP paths. Can you help me with the instructions 
how to configure my ingress LER for this purpose? Thanks in advance.

Regards,

Khac Nguyen

Re: [mpls-linux-general] How to patch the kernel and turn off debug?

[<gui...@ya...>]

Hello,

I found another way to turn off the kernel debug in
older  messages:

" In kernel source code ... include/net/mpls.h: (about
line 57)
 comment out these two #defines and the debug code
will be omited,
  as opposed to just not sent to debug logs.
 
 /* Comment this to suppress MPLS_DEBUG calls */
 #define MPLS_ENABLE_DEBUG 1
 
 /* Comment this to suppress TRACING enter/exit
functions */
 #define MPLS_ENABLE_DEBUG_FUNC 1"

Thanks anyway

__________________________________________________
Faça ligações para outros computadores com o novo Yahoo! Messenger 
http://br.beta.messenger.yahoo.com/ 

[mpls-linux-general] How to patch the kernel and turn off debug?

[<gui...@ya...>]

Hello,

I'm trying to apply a Kernel patch to turn off de
Debug option of my mpls-linux but I'm getting the
error below:

----------
[root@compute-0-15 linux]# patch -p1 <
../debugoff.diff
patching file include/net/mpls.h
Hunk #1 succeeded at 63 (offset 4 lines).
patching file net/Kconfig
Hunk #1 FAILED at 96.
1 out of 1 hunk FAILED -- saving rejects to file
net/Kconfig.rej
patching file net/mpls/Makefile
Hunk #1 FAILED at 9.
1 out of 1 hunk FAILED -- saving rejects to file
net/mpls/Makefile.rej
------------

My kernel source version is: 2.6.9
(/usr/src/linux-2.6.9) 
My Mpls version: 1.946a
My SO: FC4

I got the patch that I tryed do apply from the list:

---------------------
diff -uprN -X dontdiff kernel/include/net/mpls.h
kernel-bb/include/net/mpls.h
--- kernel/include/net/mpls.h   2005-12-07
10:08:49.000000000 -0500
+++ kernel-bb/include/net/mpls.h        2005-12-07
15:51:52.000000000 -0500
@@ -59,6 +59,15 @@ extern struct dst_ops mpls_dst_ops;
 /* Comment this to suppress TRACING enter/exit
functions */
 #define MPLS_ENABLE_DEBUG_FUNC 1

+/* Override driver debugging from Kconfig option:
+ * MPLS_DISABLE_DEBUG
+ */
+#ifdef MPLS_DISABLE_DEBUG
+#undefine MPLS_ENABLE_DEBUG
+#undefine MPLS_ENABLE_DEBUG_FUNC
+#endif
+
+
 #ifdef  MPLS_ENABLE_DEBUG
 #define MPLS_DEBUG(f, a...) \
 { \
diff -uprN -X dontdiff kernel/net/Kconfig
kernel-bb/net/Kconfig
--- kernel/net/Kconfig  2005-12-07 10:08:49.000000000
-0500
+++ kernel-bb/net/Kconfig       2005-12-07
15:53:38.000000000 -0500
@@ -96,6 +96,17 @@ config MPLS_TUNNEL

          If unsure, say N.

+config MPLS_DISABLE_DEBUG
+       boolean "MPLS: Disable Debug"
+       depends on MPLS
+       default n
+       ---help---
+         This option allows to disable the inclusion
of debugging code in
+         the MPLS driver.  The mpls.h file has the
actual DEBUG #defines,
+         but saying Y here, will force them off.
+
+         If you feel the need for speed, say Y.
+
 if INET
 source "net/ipv4/Kconfig"
 source "net/ipv6/Kconfig"
diff -uprN -X dontdiff kernel/net/mpls/Makefile
kernel-bb/net/mpls/Makefile
--- kernel/net/mpls/Makefile    2005-12-07
10:08:49.000000000 -0500
+++ kernel-bb/net/mpls/Makefile 2005-12-07
15:54:00.000000000 -0500
@@ -9,3 +9,7 @@ mpls-y := af_mpls.o mpls_if.o
mpls_ilm.o

 obj-$(CONFIG_MPLS) += mpls.o
 obj-$(CONFIG_MPLS_TUNNEL) += mpls_tunnel.o
+
+ifeq ($(CONFIG_MPLS_DISABLE_DEBUG),y)
+EXTRA_CFLAGS += -DMPLS_DISABLE_DEBUG
+endif
--------------

What can be wrong?

Best Regards






	



	
		
_______________________________________________________ 
Yahoo! doce lar. Faça do Yahoo! sua homepage. 
http://br.yahoo.com/homepageset.html 

[mpls-linux-general] please help me more

[diva p. <div...@pl...>]

Hii James, i have a trouble with my mpls hierarchy 
configuration setup.


My hierarchy is:

  

                        *---------LSRB---------*
                        |                              |
                        |                              |
HostA---------LER1--------------------LER2-----------HostB
                        |                              |
                        |                              |
                        *---------LSRB---------*


Network address:
-->HostA to LER1 : 10.14.1.0/24
-->LER1 to LSRB  : 10.14.2.0/24
-->LER1 to LSRB  : 10.14.4.0/24
-->LER1 to LER2  : 10.14.3.0/24
-->LER2 to HostB : 10.14.7.0/24


I used :
-->mpls-linux-2.6.9.tar.bz2
-->iproute-2.6.9-3mpls1_946.i386.rpm
-->kernel linux-2.6.9.tar.bz2

I confused when i used static route mpls from LER1,i 
couldn't route with 3 gateway, there are from
LER1 with LSRA as first gateway
LER1 with LER2 as second gateway
LER1 with LSRB as third gateway.

I have used static but when setting up the seond gateway 
failed with RNETLINK.You can see below
LER1
------
mpls nhlfe add key 0 instructions push gen 100 nexthop 
eth1 ipv4 10.14.2.2
(key1)
ip route add 10.14.5.0/24 via 10.14.2.2 spec_nh 0x8847 
key1  (first gateway)
mpls nhlfe add key 0 instructions push gen 200 nexthop 
eth1 ipv4 10.14.3.2
(key2)
ip route add 10.14.5.0/24 via 10.14.3.2 spec_nh 0x8847 
key2  (second gateway)
RTNETLINK:no success
---------------------------------------------------------------------------

Do I have to use dinamic route with Quagga???
Ihave compiled quagga-0.97.3-1.FC3mpls1_946a.i386 
successfully,but i ccouldn't enter to CLI router.Do I need 
other packages??

Could mpls command use priority route if the first gateway 
faild automaticaly routing move to second gateway or to 
third gateway????


Please give me some suggestion.
========================================================================================
Ikuti Lomba Puisi Online Jawa Timur, dapatkan hadiah menarik setiap bulannya 
dan hadiah total senilai 60 juta rupiah
hanya di http://www.plasa.com (Khusus Jawa Timur)
======================================================================================== 

[mpls-linux-general] please help me

[diva p. <div...@pl...>]

Hii James, i have a trouble with my mpls hierarchy 
configuration setup.


My hierarchy is:

  

                        *---------LSRB---------*
                        |                              |
                        |                              |
HostA---------LER1--------------------LER2-----------HostB
                        |                              |
                        |                              |
                        *---------LSRB---------*


Network address:
-->HostA to LER1 : 10.14.1.0/24
-->LER1 to LSRB  : 10.14.2.0/24
-->LER1 to LSRB  : 10.14.4.0/24
-->LER1 to LER2  : 10.14.3.0/24
-->LER2 to HostB : 10.14.7.0/24


I used :
-->mpls-linux-2.6.9.tar.bz2
-->iproute-2.6.9-3mpls1_946.i386.rpm
-->kernel linux-2.6.9.tar.bz2

I confused when i used static route mpls from LER1,i 
couldn't route with 3 gateway, there are from
LER1 with LSRA as first gateway
LER1 with LER2 as second gateway
LER1 with LSRB as third gateway.

I have used static but when setting up the seond gateway 
failed with RNETLINK.You can see below
LER1
------
mpls nhlfe add key 0 instructions push gen 100 nexthop 
eth1 ipv4 10.14.2.2
(key1)
ip route add 10.14.5.0/24 via 10.14.2.2 spec_nh 0x8847 
key1  (first gateway)
mpls nhlfe add key 0 instructions push gen 200 nexthop 
eth1 ipv4 10.14.3.2
(key2)
ip route add 10.14.5.0/24 via 10.14.3.2 spec_nh 0x8847 
key2  (second gateway)
RTNETLINK:no success
---------------------------------------------------------------------------

Do I have to use dinamic route with Quagga???
Ihave compiled quagga-0.97.3-1.FC3mpls1_946a.i386 
successfully,but i ccouldn't enter to CLI router.Do I need 
other packages??

Could mpls command use priority route if the first gateway 
faild automaticaly routing move to second gateway or to 
third gateway????


Please give me some suggestion.
========================================================================================
Ikuti Lomba Puisi Online Jawa Timur, dapatkan hadiah menarik setiap bulannya 
dan hadiah total senilai 60 juta rupiah
hanya di http://www.plasa.com (Khusus Jawa Timur)
======================================================================================== 

Re: [mpls-linux-general] regarding project on mpls-linux

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

Take a look at the code in iproute2-mpls (in my development tree or in the
iproute SRPM).  Specifically the code that makes up the 'mpls' utility.  It
shows how to go about creating and retrieving info about the MPLS forwarding
plane.

You may also want to look at:

    http://sourceforge.net/projects/mpls-agent

On Wed, Feb 15, 2006 at 01:12:59PM +0530, kis...@wi... wrote:
> Hii
> =20
>          I am developing a snmp agent (having MIB for QOS parameters like=
 delay, jitter, latency etc as objects) interfacing with mpls-linux driver.=
 I am working in FC 3. I would like to know how do I access the driver info=
rmation from my application?. How do I get label info , Label table? Is the=
re any functions used?.. How to proceed with my project as I am new to the =
environment. Please guide me.
> =20
> =20
> Regards,
> Kishor

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] RE: regarding project on mpls-linux

[<kis...@wi...>]

=20
=20
-----Original Message-----
From: Kishor Krishna=20
Sent: Wednesday, February 15, 2006 1:13 PM
To: 'mpl...@li...'
Subject: regarding project on mpls-linux
=20
Hii
=20
         I am developing a snmp agent (having MIB for QOS parameters =
like delay, jitter, latency etc as objects) interfacing with mpls-linux =
driver. I am working in FC 3. I would like to know how do I access the =
driver information from my application?. How do I get LSP creation from =
source to destination, bandwidth, delay etc...How do I get label info, =
Label table? Is there any functions used?.. How to proceed with my =
project as I am new to the environment. Please guide me.
=20
=20
Regards,
Kishor

[mpls-linux-general] regarding project on mpls-linux

[<kis...@wi...>]

Hii
=20
         I am developing a snmp agent (having MIB for QOS parameters =
like delay, jitter, latency etc as objects) interfacing with mpls-linux =
driver. I am working in FC 3. I would like to know how do I access the =
driver information from my application?. How do I get label info , Label =
table? Is there any functions used?.. How to proceed with my project as =
I am new to the environment. Please guide me.
=20
=20
Regards,
Kishor

Re: [mpls-linux-general] How to measure traffic in the core of a MPLS network

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

Hmm.  This is a good question.

So first I want to point out that are counter stored on each ILM and
NHLFE.  You can get access to these counters via netlink.

So I do not know of a way to directly match the label.  This sounds
like an excellent opportunity for a project if anyone is looking for one.
I can already image ways of doing this. (mplstables anyone?)

I don't think iptables is going to help you at all in this effort
since it only can interact with packets that go into the IPv4
stack.  In the core of the network the packet never leave the MPLS
stack.

TC might be able to help by using its 'match' code which allows you
to specify an offset and a hex-value to match against (pretty tedious).
Once you have a match I think you could apply some "tc actions" to the
packet.  I'm not sure what all actions you can apply, but I think there
is a mechanism for counting packets.

Let me know what you find.

On Mon, Feb 13, 2006 at 10:59:38PM +0000, Guilherme Torres wrote:
> I'm trying to measure the MPLS traffic between two
> LSR's in the core of a MPLS network. How do I do to
> match the label of a MPLS packet in a iptables rule?=20
>=20
> I tryed to use a rule like "iptables -A FORWARD -m mac
> --mac-source 00:A0:C9:73:EF:77 -j ACCEPT" but it
> doesn't count any byte.
>=20
> Best regards
>=20
>=20
> 	=09
> _______________________________________________________
> Yahoo! Acesso Gr?tis - Internet r?pida e gr?tis. Instale o discador agora!
> http://br.acesso.yahoo.com
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=
=3D121642
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] How to measure traffic in the core of a MPLS network

[Guilherme T. <gui...@ya...>]

I'm trying to measure the MPLS traffic between two
LSR's in the core of a MPLS network. How do I do to
match the label of a MPLS packet in a iptables rule? 

I tryed to use a rule like "iptables -A FORWARD -m mac
--mac-source 00:A0:C9:73:EF:77 -j ACCEPT" but it
doesn't count any byte.

Best regards


		
_______________________________________________________
Yahoo! Acesso Grátis - Internet rápida e grátis. Instale o discador agora!
http://br.acesso.yahoo.com

Re: [mpls-linux-general] Question on MPLS linux

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

Version 1.950 has forwarding plane support for Ethernet over MPLS.

On Sun, Feb 12, 2006 at 03:16:07AM +0900, ????????? wrote:
> Dear R. Leu.
>=20
> The current mpls linux supports EoMPLS (ethernet over MPLS)?
>=20
> -jeongho

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] Question on MPLS linux

[<jeo...@co...>]

RGVhciBSLiBMZXUuDQoNClRoZSBjdXJyZW50IG1wbHMgbGludXggc3VwcG9ydHMgRW9NUExTIChl
dGhlcm5ldCBvdmVyIE1QTFMpPw0KDQotamVvbmdobw==

Re: [mpls-linux-general] configuration FC3 mpls successfully but trouble with bandwidth, please help me

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

This has been talked to death on the mailing already. Read in the archive
for details, but:

- if your using the RPMs I distribute, then your using a debug build of
  the kernel that generates TONS of messages, each printk is incredibly slow
- in a debug build you can turn off generation of debug messages, but that
  does not prevent all of the debug code from running
- Egrees LERs will ALWAYS be slower then an IPv4 only device:  MPLS lookup
  and IPv4 lookup
- higher latency does not equal slower

On Sat, Feb 11, 2006 at 10:57:45PM +0700, diva permata wrote:
>=20
> Hii James, i need your help please.
>=20
> I have installed mpls-linux FC3 succesfully.With the=20
> hierarchy below:
>=20
>=20
>=20
>=20
> I used :
> -->mpls-linux-2.6.9.tar.bz2
> -->iproute-2.6.9-3mpls1_946.i386.rpm
> -->kernel linux-2.6.9.tar.bz2
> -->used Fast Ethernet 100Mbps in all interfaces
>=20
>                                                           =20
>      101
>                                      100=20
>                      201                        200
> HostA---(eth0)LER1(eth1)----(eth0)LSR1(eth1)----(eth0)LSR2(eth1)----(eth0=
)LER2(eth1)----HostB
>=20
>=20
> I have installed FTP server in Host B.And when i got data=20
> from HostB with FTP client in hostA,i got some trouble why=20
> ip-based is faster than mpls-based. The data is below:
>=20
> size                     ip-based(delay)=20
>             mpls-based(delay)
> -----                    ------------------=20
>             ---------------------
> 1Mb                      0,299 sec=20
>                         3,77 sec
> 2Mb                       0,235 sec=20
>                        7,51 sec
> 4Mb                       0,436 sec=20
>                        15 sec
> 6Mb                       0,677 sec=20
>                        22,4 sec
> 8Mb                       0,677 sec=20
>                        29,9 sec
> 9,5Mb                    1 sec=20
>                               34,6 sec
>=20
> I have check all of cables everything is oke. Is the=20
> output normal,MR.James????
> And then when i 'ping' from hostA to hostB have time=3Dabout=20
> 3-5 bit/sec
>=20
> My Configuration is:
> ----------------------
> LER1
> ------
> mpls nhlfe add key 0 instructions push gen 100 nexthop=20
> eth1 ipv4 10.14.2.2
> ip route add 10.14.5.0/24 via 10.14.2.2 spec_nh 0x8847 0x2
> mpls labelspace add dev eth1 labelspace 0
> mpls ilm add label gen 100 labelspace 0
>=20
> LSR1
> ------
> mpls labelspace add dev eth0 labelspace 0
> mpls ilm add label gen 100 labelspace 0
> mpls nhlfe add key 0 instructions push gen 101 nexthop=20
> eth1 ipv4 10.14.3.2
> mpls xc add ilm_label gen 100 ilm_labelspace 0 nhlfe_key=20
> 0x2
> mpls labelspace add dev eth1 labelspace 0
> mpls ilm add label gen 201 labelspace 0
> mpls nhlfe add key 0 instructions push gen 100 nexthop=20
> eth0 ipv4 10.14.2.1
> mpls xc add ilm_label gen 201 ilm_labelspace 0 nhlfe_key=20
> 0x3
>=20
> LSR2
> ------
> mpls labelspace add dev eth0 labelspace 0
> mpls ilm add label gen 101 labelspace 0
> mpls nhlfe add key 0 instructions push gen 200 nexthop=20
> eth1 ipv4 10.14.4.2
> mpls xc add ilm_label gen 101 ilm_labelspace 0 nhlfe_key=20
> 0x2
> mpls labelspace add dev eth1 labelspace 0
> mpls ilm add label gen 200 labelspace 0
> mpls nhlfe add key 0 instructions push gen 201 nexthop=20
> eth0 ipv4 10.14.3.1
> mpls xc add ilm_label gen 200 ilm_labelspace 0 nhlfe_key=20
> 0x3
>=20
> LER2
> -----
> mpls labelspace add dev eth0 labelspace 0
> mpls ilm add label gen 200 labelspace 0
> mpls nhlfe add key 0 instructions push gen 200 nexthop=20
> eth0 ipv4 10.14.4.1
> ip route add 10.14.1.0/24 via 10.14.4.1 spec_nh 0x8847 0x2
>=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=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
> Ikuti Lomba Puisi Online Jawa Timur, dapatkan hadiah menarik setiap=20
> bulannya dan hadiah total senilai 60 juta rupiah
> hanya di http://www.plasa.com (Khusus Jawa Timur)
> =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=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=20
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=
=3D121642
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

--=20
James R. Leu
jl...@mi...

[mpls-linux-general] configuration FC3 mpls successfully but trouble with bandwidth, please help me

[diva p. <div...@pl...>]

Hii James, i need your help please.

I have installed mpls-linux FC3 succesfully.With the 
hierarchy below:




I used :
-->mpls-linux-2.6.9.tar.bz2
-->iproute-2.6.9-3mpls1_946.i386.rpm
-->kernel linux-2.6.9.tar.bz2
-->used Fast Ethernet 100Mbps in all interfaces

                                                            
      101
                                      100 
                      201                        200
HostA---(eth0)LER1(eth1)----(eth0)LSR1(eth1)----(eth0)LSR2(eth1)----(eth0)LER2(eth1)----HostB


I have installed FTP server in Host B.And when i got data 
from HostB with FTP client in hostA,i got some trouble why 
ip-based is faster than mpls-based. The data is below:

size                     ip-based(delay) 
             mpls-based(delay)
-----                    ------------------ 
             ---------------------
1Mb                      0,299 sec 
                         3,77 sec
2Mb                       0,235 sec 
                        7,51 sec
4Mb                       0,436 sec 
                        15 sec
6Mb                       0,677 sec 
                        22,4 sec
8Mb                       0,677 sec 
                        29,9 sec
9,5Mb                    1 sec 
                               34,6 sec

I have check all of cables everything is oke. Is the 
output normal,MR.James????
And then when i 'ping' from hostA to hostB have time=about 
3-5 bit/sec

My Configuration is:
----------------------
LER1
------
mpls nhlfe add key 0 instructions push gen 100 nexthop 
eth1 ipv4 10.14.2.2
ip route add 10.14.5.0/24 via 10.14.2.2 spec_nh 0x8847 0x2
mpls labelspace add dev eth1 labelspace 0
mpls ilm add label gen 100 labelspace 0

LSR1
------
mpls labelspace add dev eth0 labelspace 0
mpls ilm add label gen 100 labelspace 0
mpls nhlfe add key 0 instructions push gen 101 nexthop 
eth1 ipv4 10.14.3.2
mpls xc add ilm_label gen 100 ilm_labelspace 0 nhlfe_key 
0x2
mpls labelspace add dev eth1 labelspace 0
mpls ilm add label gen 201 labelspace 0
mpls nhlfe add key 0 instructions push gen 100 nexthop 
eth0 ipv4 10.14.2.1
mpls xc add ilm_label gen 201 ilm_labelspace 0 nhlfe_key 
0x3

LSR2
------
mpls labelspace add dev eth0 labelspace 0
mpls ilm add label gen 101 labelspace 0
mpls nhlfe add key 0 instructions push gen 200 nexthop 
eth1 ipv4 10.14.4.2
mpls xc add ilm_label gen 101 ilm_labelspace 0 nhlfe_key 
0x2
mpls labelspace add dev eth1 labelspace 0
mpls ilm add label gen 200 labelspace 0
mpls nhlfe add key 0 instructions push gen 201 nexthop 
eth0 ipv4 10.14.3.1
mpls xc add ilm_label gen 200 ilm_labelspace 0 nhlfe_key 
0x3

LER2
-----
mpls labelspace add dev eth0 labelspace 0
mpls ilm add label gen 200 labelspace 0
mpls nhlfe add key 0 instructions push gen 200 nexthop 
eth0 ipv4 10.14.4.1
ip route add 10.14.1.0/24 via 10.14.4.1 spec_nh 0x8847 0x2




========================================================================================
Ikuti Lomba Puisi Online Jawa Timur, dapatkan hadiah menarik setiap bulannya 
dan hadiah total senilai 60 juta rupiah
hanya di http://www.plasa.com (Khusus Jawa Timur)
========================================================================================