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

Re: [mpls-linux-general] Where does it start?

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

Since you've install the mplsadm2 RPM, now the mplsadm2 binary is installed
in /usr/sbin/.  Make sure /usr/sbin/ is in your PATH.  If so then you can
run mplsadm2 from anywhere.

On Thu, Nov 04, 2004 at 05:48:33PM +0100, windtim@libero.i wrote:
> Hi=20
>    i've installed listed software and i'm going to start my test but... w=
here i
> can do it? In past release (mpls-1.119) i use directory /usr/src/mpls-lin=
ux1.1
> (or something similar) to give my instructions. Now? Where  do i start?
>=20
>=20
>=20
> ____________________________________________________________
> Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazio=
ne.=20
> Abbonati subito su http://www.libero.it=20
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_idU88&alloc_id=12065&op?k
> _______________________________________________
> 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] My first LSP (Strange output)

[windtim\@libero\.i <wi...@li...>]

Hi 
 i've read the example on http://mpls-linux.sourceforge.net/ and try=
ing to copy
i've the followe output:
[root@localhost root]# mplsadm2 -A=
 -O 0
MPLSADM mplsadm.c:692:main: Out segment input: 0
Key: 0x00000002=0D
=
Out Segment add: Success
[root@localhost root]#
What's happened? Is thi=
s ok or something goes wrong? In my opinion i don't have
nothing to care=
 but the first message isn't usual in my old test
=0A=0A=0A=0A__________=
__________________________________________________=0ALibero ADSL: navighi=
 gratis a 1.2 Mega, senza canone e costi di attivazione. =0AAbbonati subi=
to su http://www.libero.it =0A

[mpls-linux-general] Where does it start?

[windtim\@libero\.i <wi...@li...>]

Hi 
   i've installed listed software and i'm going to start my test but=
... where i
can do it? In past release (mpls-1.119) i use directory /usr=
/src/mpls-linux1.1
(or something similar) to give my instructions. Now? =
Where  do i start?=0A=0A=0A=0A___________________________________________=
_________________=0ALibero ADSL: navighi gratis a 1.2 Mega, senza canone =
e costi di attivazione. =0AAbbonati subito su http://www.libero.it =0A

Re: [mpls-linux-general] Questions about rpms for MPLS

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

On Thu, Nov 04, 2004 at 11:18:04AM +0100, windtim@libero.i wrote:
> Hi,
>      i've downloaded software suggested for MPLS for Linux from the URL
> indicated. My question is: is linux-kernel...mpls.rpm compiled for a part=
icular
> architecture (i386, i684, ia_64 or other ones) or i can install on my Cen=
trino
> 1.6 Ghz and all other pcs? I'd like to know it before install it.
> I've linux-kernel-2.6.5 and, on James's suggest, i'd like to install
> linux..2.6.8..rpm without download linux-kernel.2.6.8 before
> Thank in advance for the help

The RPMS I build work on all i686 compatibles ... including Centrinos.
You do not need to download and compile your own kernel if you use my
RPMs ... that is why I create them, so you don't have to :-)

>=20
>=20
>=20
> ____________________________________________________________
> Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazio=
ne.=20
> Abbonati subito su http://www.libero.it=20
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_idU88&alloc_id=12065&op?k
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

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

Re: [mpls-linux-general] latest mpls-kernel-1.1 (2.6.9)

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

I have pending changes that fix this.  I have submitted them (change 994)

On Thu, Nov 04, 2004 at 07:51:55AM +0200, Peter Shtinkov wrote:
> Hello,
> 
> Is there any requirements for gcc compiler or something in
> mpls-kernel-1.1 from perforce. I synced with Change 992, and there are
> no more changes currently available in kernel.
> 
> It is 2.6.9 based but connot be compiled with this:
> 
> Reading specs from /usr/lib/gcc-lib/i486-linux/3.3.5/specs
> Configured with: ../src/configure -v --enable-languages=c,c
> ++,java,f77,pascal,objc,ada,treelang --prefix=/usr
> --mandir=/usr/share/man --infodir=/usr/share/info
> --with-gxx-include-dir=/usr/include/c++/3.3 --enable-shared
> --with-system-zlib --enable-nls --without-included-gettext
> --enable-__cxa_atexit --enable-clocale=gnu --enable-debug
> --enable-java-gc=boehm --enable-java-awt=xlib --enable-objc-gc
> i486-linux
> Thread model: posix
> gcc version 3.3.5 (Debian 1:3.3.5-2)
> 
> Here are the errors:
> 
>  CC      net/core/gen_estimator.o
> In file included from net/core/gen_estimator.c:35:
> include/net/gen_stats.h:30: warning: `struct gnet_stats_basic' declared
> inside parameter list
> include/net/gen_stats.h:30: warning: its scope is only this definition
> or declaration, which is probably not what you want
> include/net/gen_stats.h:32: warning: `struct gnet_stats_rate_est'
> declared inside parameter list
> include/net/gen_stats.h:34: warning: `struct gnet_stats_queue' declared
> inside parameter list
> include/net/gen_stats.h:41: warning: `struct gnet_stats_rate_est'
> declared inside parameter list
> include/net/gen_stats.h:41: warning: `struct gnet_stats_basic' declared
> inside parameter list
> include/net/gen_stats.h:43: warning: `struct gnet_stats_rate_est'
> declared inside parameter list
> include/net/gen_stats.h:43: warning: `struct gnet_stats_basic' declared
> inside parameter list
> net/core/gen_estimator.c: In function `est_timer':
> net/core/gen_estimator.c:117: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c:118: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c:122: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c:127: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c: At top level:
> net/core/gen_estimator.c:137: error: conflicting types for
> `gen_new_estimator'
> include/net/gen_stats.h:41: error: previous declaration of
> `gen_new_estimator'
> net/core/gen_estimator.c: In function `gen_new_estimator':
> net/core/gen_estimator.c:157: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c:158: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c:159: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c:160: error: dereferencing pointer to incomplete
> type
> net/core/gen_estimator.c: At top level:
> net/core/gen_estimator.c:178: error: conflicting types for
> `gen_kill_estimator'
> include/net/gen_stats.h:43: error: previous declaration of
> `gen_kill_estimator'
> make[3]: *** [net/core/gen_estimator.o] Error 1
> make[2]: *** [net/core] Error 2
> make[1]: *** [net] Error 2
> make[1]: Leaving directory `/usr/src/mpls-kernel-1.1'
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

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

[mpls-linux-general] Questions about rpms for MPLS

[windtim\@libero\.i <wi...@li...>]

Hi,
     i've downloaded software suggested for MPLS for Linux from the =
URL
indicated. My question is: is linux-kernel...mpls.rpm compiled for a=
 particular
architecture (i386, i684, ia_64 or other ones) or i can inst=
all on my Centrino
1.6 Ghz and all other pcs? I'd like to know it before=
 install it.
I've linux-kernel-2.6.5 and, on James's suggest, i'd like t=
o install
linux..2.6.8..rpm without download linux-kernel.2.6.8 before=0D
=
Thank in advance for the help=0A=0A=0A=0A________________________________=
____________________________=0ALibero ADSL: navighi gratis a 1.2 Mega, se=
nza canone e costi di attivazione. =0AAbbonati subito su http://www.liber=
o.it =0A

[mpls-linux-general] latest mpls-kernel-1.1 (2.6.9)

[Peter S. <sht...@sp...>]

Hello,

Is there any requirements for gcc compiler or something in
mpls-kernel-1.1 from perforce. I synced with Change 992, and there are
no more changes currently available in kernel.

It is 2.6.9 based but connot be compiled with this:

Reading specs from /usr/lib/gcc-lib/i486-linux/3.3.5/specs
Configured with: ../src/configure -v --enable-languages=c,c
++,java,f77,pascal,objc,ada,treelang --prefix=/usr
--mandir=/usr/share/man --infodir=/usr/share/info
--with-gxx-include-dir=/usr/include/c++/3.3 --enable-shared
--with-system-zlib --enable-nls --without-included-gettext
--enable-__cxa_atexit --enable-clocale=gnu --enable-debug
--enable-java-gc=boehm --enable-java-awt=xlib --enable-objc-gc
i486-linux
Thread model: posix
gcc version 3.3.5 (Debian 1:3.3.5-2)

Here are the errors:

 CC      net/core/gen_estimator.o
In file included from net/core/gen_estimator.c:35:
include/net/gen_stats.h:30: warning: `struct gnet_stats_basic' declared
inside parameter list
include/net/gen_stats.h:30: warning: its scope is only this definition
or declaration, which is probably not what you want
include/net/gen_stats.h:32: warning: `struct gnet_stats_rate_est'
declared inside parameter list
include/net/gen_stats.h:34: warning: `struct gnet_stats_queue' declared
inside parameter list
include/net/gen_stats.h:41: warning: `struct gnet_stats_rate_est'
declared inside parameter list
include/net/gen_stats.h:41: warning: `struct gnet_stats_basic' declared
inside parameter list
include/net/gen_stats.h:43: warning: `struct gnet_stats_rate_est'
declared inside parameter list
include/net/gen_stats.h:43: warning: `struct gnet_stats_basic' declared
inside parameter list
net/core/gen_estimator.c: In function `est_timer':
net/core/gen_estimator.c:117: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c:118: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c:122: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c:127: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c: At top level:
net/core/gen_estimator.c:137: error: conflicting types for
`gen_new_estimator'
include/net/gen_stats.h:41: error: previous declaration of
`gen_new_estimator'
net/core/gen_estimator.c: In function `gen_new_estimator':
net/core/gen_estimator.c:157: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c:158: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c:159: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c:160: error: dereferencing pointer to incomplete
type
net/core/gen_estimator.c: At top level:
net/core/gen_estimator.c:178: error: conflicting types for
`gen_kill_estimator'
include/net/gen_stats.h:43: error: previous declaration of
`gen_kill_estimator'
make[3]: *** [net/core/gen_estimator.o] Error 1
make[2]: *** [net/core] Error 2
make[1]: *** [net] Error 2
make[1]: Leaving directory `/usr/src/mpls-kernel-1.1'

Re: [mpls-linux-general] Help me on installation

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

On Wed, Nov 03, 2004 at 04:53:07PM +0100, windtim@libero.i wrote:
> Hi everybody,
>         i've Fedora Core 2 with Linux-kernel-2.6.5-8. Is there anyone tha=
t has
> made his test on MPLS for Linux with this kernel release? How does it fee=
l? Is
> there any problem? Which one? Explain me clearly please. Thanks in advanc=
e for
> the help
> P.S.: If i have to use linux kernel 2.6.8 can i download directly
> linux..kernel..mpls..rpm and rpm out it or i've to download before linux =
kernel

You can use the 2.6.8.1 kernel RPM I have posted in the file section
of http://sf.net/projects/mpls-linux/

One note, you will probably want to use 'rpm -ivh' not 'rpm -Uvh'
this will enable you to have 2 kernels installed on the system
so if for some reason the MPLS kernel doesn't work, you can always
boot the old kernel.

You will also need the other RPMS (iproute2 iptables mplssadm2).
You should use 'rpm -Uvh' when install these RPMs.


> 2.6.8 and after "upgrade" it?
> Thanks a lot yet
>=20
>=20
>=20
> ____________________________________________________________
> Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazio=
ne.=20
> Abbonati subito su http://www.libero.it=20
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_idU88&alloc_id=12065&op?k
> _______________________________________________
> 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] Help me on installation

[windtim\@libero\.i <wi...@li...>]

Hi everybody,
        i've Fedora Core 2 with Linux-kernel-2.6.5-8. Is t=
here anyone that has
made his test on MPLS for Linux with this kernel re=
lease? How does it feel? Is
there any problem? Which one? Explain me cle=
arly please. Thanks in advance for
the help
P.S.: If i have to use linu=
x kernel 2.6.8 can i download directly
linux..kernel..mpls..rpm and rpm =
out it or i've to download before linux kernel
2.6.8 and after "upgrade"=
 it?
Thanks a lot yet=0A=0A=0A=0A_______________________________________=
_____________________=0ALibero ADSL: navighi gratis a 1.2 Mega, senza can=
one e costi di attivazione. =0AAbbonati subito su http://www.libero.it =0A=

Re: [mpls-linux-general] working od ldpd

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

On Tue, Nov 02, 2004 at 06:14:05AM -0000, laxman quagga wrote:
> 
>  Hi,
> 
>   I compiled ldp-portable with quagga-0.96.4.
> 
>  i am not using the mpls-linux kernel
> 
>   Now I have  zebra,ospfd and ldpd daemons.
> 
>  I am able to run  daemons in the following order.
> 
>  1) Zebra
>   
>  2) ospfd
> 
>  3) ldpd.
> 
>  even it is writing something into the specified log files.
> 
> But I wanted to know the working of ldpd for better understanding.

Read the source.  I have used the Appendix A procedures and and naming
scheme from RFC 3036.

> will this ldpd do the label distribution?

Yes. But if you are not running on a mpls-linux kernel ldpd will not
install the labels into the forwarding plane.

>  help me
> 
> thanks and regards
> 
>  laxman
-- 
James R. Leu
jl...@mi...

[mpls-linux-general] working od ldpd

[laxman q. <qua...@re...>]

=0A Hi,=0A=0A  I compiled ldp-portable with quagga-0.96.4.=0A=0A i am not u=
sing the mpls-linux kernel=0A=0A  Now I have  zebra,ospfd and ldpd daemons.=
=0A=0A I am able to run  daemons in the following order.=0A=0A 1) Zebra=0A =
 =0A 2) ospfd=0A=0A 3) ldpd.=0A=0A even it is writing something into the sp=
ecified log files.=0A=0ABut I wanted to know the working of ldpd for better=
 understanding.=0A=0Awill this ldpd do the label distribution?=0A=0A=0A hel=
p me=0A=0Athanks and regards=0A=0A laxman

[mpls-linux-general] Re: mpls-linux crash system after mplsadm -A -I gen:**:0

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

On Fri, Oct 29, 2004 at 08:16:03PM +0000, Christian Beier wrote:
> James R. Leu wrote:
> 
> >If you can't complile a 2.6.8.1 kernel, what kernel do did the above 
> >'crash'
> >occur on?
> SuSE's 2.6.4-52-default kernel but I didn't believe in any kernel-prob,
> cause both boxes had have the same, one is working, one is not. Of
> course the working one uses mpls-linux-1.935.
> The problematic one crashed with SuSE's 2.6.4-52 and nearly same
> kernel-config (except smp) while "mplsadm2 -A  -I gen:*:0" so i updated
> to mpls-linux-1.938 and after that, there was an output-error after
> "mplsadm2 -A -O 0" so i tried tried it with kernel 2.6.8.1 .
> >Don't compile MPLS or any of it's componets as modules.  There are
> >fixes pending for this.
> Jepp...I've read it also and didn't do it, just wrong discription. It
> was while making of one "net/mpls*.o" files...some related with arp, i
> think.
> I could say it clearly when I fixed the "fib_semantic.o" problem, I fear.
> 
> >
> >>"...
> >> CC      net/ipv4/fib_semantics.o
> >>net/ipv4/fib_semantics.c: In function `fib_dump_info':
> >>net/ipv4/fib_semantics.c:719: error: `RTA_SPEC_PROTO' undeclared (first 
> >>use in this function)
> >>net/ipv4/fib_semantics.c:719: error: (Each undeclared identifier is 
> >>reported only once
> >>net/ipv4/fib_semantics.c:719: error: for each function it appears in.)
> >>net/ipv4/fib_semantics.c:721: error: `RTA_SPEC_DATA' undeclared (first 
> >>use in this function)
> >>make[2]: *** [net/ipv4/fib_semantics.o] Error 1
> >>make[1]: *** [net/ipv4] Error 2
> >>make: *** [net] Error 2
> >
> >
> >You're linux/include/linux/rtnetlink.h was not update by the
> >pactch.  Check for '*.rej' in your kernel tree.
> There is a Makefile.rej.
> mmmh...stupid question of an newb...how to fix it manually?

You have to look a the *.rej file and try and manually edit the file.

> Thx 4fast help
> Christian
> 
> @ James:
> 	Please forward it again and check my account! I may receive mails 
> 	but couldn't send to list address.

Make sure your mail client is setting the from address to the
same address you registered with the mailing list.

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

[mpls-linux-general] Re: mpls-linux crash system after mplsadm -A -I gen:**:0

[Christian B. <chr...@st...>]

James R. Leu wrote:

> If you can't complile a 2.6.8.1 kernel, what kernel do did the above 'crash'
> occur on?
SuSE's 2.6.4-52-default kernel but I didn't believe in any kernel-prob,
cause both boxes had have the same, one is working, one is not. Of
course the working one uses mpls-linux-1.935.
The problematic one crashed with SuSE's 2.6.4-52 and nearly same
kernel-config (except smp) while "mplsadm2 -A  -I gen:*:0" so i updated
to mpls-linux-1.938 and after that, there was an output-error after
"mplsadm2 -A -O 0" so i tried tried it with kernel 2.6.8.1 .
> Don't compile MPLS or any of it's componets as modules.  There are
> fixes pending for this.
Jepp...I've read it also and didn't do it, just wrong discription. It
was while making of one "net/mpls*.o" files...some related with arp, i
think.
I could say it clearly when I fixed the "fib_semantic.o" problem, I fear.

> 
>>"...
>>  CC      net/ipv4/fib_semantics.o
>>net/ipv4/fib_semantics.c: In function `fib_dump_info':
>>net/ipv4/fib_semantics.c:719: error: `RTA_SPEC_PROTO' undeclared (first use in this function)
>>net/ipv4/fib_semantics.c:719: error: (Each undeclared identifier is reported only once
>>net/ipv4/fib_semantics.c:719: error: for each function it appears in.)
>>net/ipv4/fib_semantics.c:721: error: `RTA_SPEC_DATA' undeclared (first use in this function)
>>make[2]: *** [net/ipv4/fib_semantics.o] Error 1
>>make[1]: *** [net/ipv4] Error 2
>>make: *** [net] Error 2
> 
> 
> You're linux/include/linux/rtnetlink.h was not update by the
> pactch.  Check for '*.rej' in your kernel tree.
There is a Makefile.rej.
mmmh...stupid question of an newb...how to fix it manually?

Thx 4fast help
Christian

@ James:
	Please forward it again and check my account! I may receive mails but 
couldn't send to list address.

Re: [mpls-linux-general] Status of MPLS for Linux

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

On Fri, Oct 29, 2004 at 10:19:08AM +0200, windtim@libero.i wrote:
> Hi everybody,
>                     i've done tests with MPLS for Linux using Linux kernel
> 2.4.20-8, now reading mails i've seen that there is only support for2.6.8. I'd
> like to know if tests with this release goes well. I've asked it because i've
> read that there were problems with kernel 2.6.7 and 2.6.8.
> Thanks in advance for the help.
> P.S.: Has someone used Red Hat 9 in his tests? How does it seem? Are there any
> problems?

I run my home network over LSPs created on 2.6.8(1) systems.  It is stable.
I would suggest upgrading to Fedora 2 and using the RPMs I've created.
You can find them under the file section at http://mpls-linux.sf.net/

> 
> 
> 
> ____________________________________________________________
> Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. 
> Abbonati subito su http://www.libero.it 
> 
> 
> 
> 
> -------------------------------------------------------
> This Newsletter Sponsored by: Macrovision 
> For reliable Linux application installations, use the industry's leading
> setup authoring tool, InstallShield X. Learn more and evaluate 
> today. http://clk.atdmt.com/MSI/go/ins0030000001msi/direct/01/
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

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

[mpls-linux-general] Status of MPLS for Linux

[windtim\@libero\.i <wi...@li...>]

Hi everybody,
                    i've done tests with MPLS for Linux us=
ing Linux kernel
2.4.20-8, now reading mails i've seen that there is onl=
y support for2.6.8. I'd
like to know if tests with this release goes wel=
l. I've asked it because i've
read that there were problems with kernel =
2.6.7 and 2.6.8.
Thanks in advance for the help.
P.S.: Has someone used=
 Red Hat 9 in his tests? How does it seem? Are there any
problems?=0A=0A=
=0A=0A____________________________________________________________=0ALibe=
ro ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. =
=0AAbbonati subito su http://www.libero.it =0A

Re: [mpls-linux-general] Router Performance

[Arturo N. <jua...@ic...>]

Hello:

On Wed, Oct 27, 2004 at 11:17:54AM -0500, James R. Leu wrote:
> Couple things to note:
> -You will never see a speed improvement on a LER<->LSR<->LER test
>  compared to normal IPv4.
> -You should see speed improvemtns for Many LERs connected to one LSR.
> -Waw label switching should be faster then IPv4 routing (you would need
>  a traffic generator that could create labeled packets.)
>

Any advice in such a type pf traffic generators ? 

Thank you,

Best Regards,
-- 
---------------------------------------------------------------------
J. Arturo Nunez                       | " When there are no role 
Intelligent Systems and Networks      | models for you to emulate,
Electrical and Electronic Engineering | create your own role and
Imperial College                      | fill it "
London UK      	                      | Author unknown
+44 (20) 7594 6218 (O)		      | 
---------------------------------------------------------------------

Re: [mpls-linux-general] Router Performance

[<bla...@in...>]

> On Fri, Oct 29, 2004 at 10:08:47AM +0700, bla...@in... wrot=
e:
>>
>> I want to test performance MPLS linux router compare with non MPLS lin=
ux
>> router.
>> does anyone ever test it??
>> give some advise pls ;)
>
> There were some results posted to this list in the past.  No performanc=
e
> testing has been done on the newer 1.9xx version which have significant
> changes that will affect speed.
>
> Couple things to note:
> -You will never see a speed improvement on a LER<->LSR<->LER test
>  compared to normal IPv4.
> -You should see speed improvemtns for Many LERs connected to one LSR.
> -Waw label switching should be faster then IPv4 routing (you would need
>  a traffic generator that could create labeled packets.)
>
>>
>>
btw, i had a plan to test it on user mode linux, will this works??
and i'm sorry first, coz my few grammar:(
on third point, what the meaning "Waw" ??

Re: [mpls-linux-general] About ldp-portable 0.800 and kernel 2.4.x

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

If you're not going to dig into the code and help fix bugs, it's probably not
read for your use.

On Wed, Oct 27, 2004 at 07:04:36PM +0200, Arturo Pablo Gonz?ffffe1lez Herranz wrote:
> Hi all again,
>            James, what do you exactly mean with "One
> thing to note is that this LDP implementation is not
> ready for users"
>            Thanks, 
>                  Art
> 
>  --- "James R. Leu" <jl...@mi...> escribi?: 
> > On Wed, Oct 27, 2004 at 10:08:35AM +0200, Arturo
> > Pablo Gonz?ffffe1lez Herranz wrote:
> > > Hi all,
> > >      I?d like to make a simple question. Could I
> > use ldp-portable 0.800 with a kernel 2.4.x patched?
> > I have to work with kernel 2.4.24 but ldp-portable
> > 0.310 doesn?t work rigth, so I?m having some
> > problems,....
> > 
> > You could make it work with some hacking, but it is
> > probably not worth the
> > effort.  I'd recomend trying it on a 2.6 kernel
> > instead.  One thing to note
> > is that this LDP implementation is not ready for
> > users.
> > 
> > >               Thanks for your time,
> > >                                            Art
> > > 
> > > 		
> > > ---------------------------------
> > > 
> > 
> > -- 
> > James R. Leu
> > jl...@mi...
> > 
> > 
> >
> -------------------------------------------------------
> > This SF.Net email is sponsored by:
> > Sybase ASE Linux Express Edition - download now for
> > FREE
> > LinuxWorld Reader's Choice Award Winner for best
> > database on Linux.
> >
> http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
> > _______________________________________________
> > mpls-linux-general mailing list
> > mpl...@li...
> >
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
> >  
> 
> 
> 		
> ______________________________________________
> Renovamos el Correo Yahoo!: ?100 MB GRATIS!
> Nuevos servicios, m?s seguridad
> http://correo.yahoo.es
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
> _______________________________________________
> 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] About ldp-portable 0.800 and kernel 2.4.x

[<apg...@ya...>]

Hi all again,
           James, what do you exactly mean with "One
thing to note is that this LDP implementation is not
ready for users"
           Thanks, 
                 Art

 --- "James R. Leu" <jl...@mi...> escribió: 
> On Wed, Oct 27, 2004 at 10:08:35AM +0200, Arturo
> Pablo Gonz?ffffe1lez Herranz wrote:
> > Hi all,
> >      I?d like to make a simple question. Could I
> use ldp-portable 0.800 with a kernel 2.4.x patched?
> I have to work with kernel 2.4.24 but ldp-portable
> 0.310 doesn?t work rigth, so I?m having some
> problems,....
> 
> You could make it work with some hacking, but it is
> probably not worth the
> effort.  I'd recomend trying it on a 2.6 kernel
> instead.  One thing to note
> is that this LDP implementation is not ready for
> users.
> 
> >               Thanks for your time,
> >                                            Art
> > 
> > 		
> > ---------------------------------
> > 
> 
> -- 
> James R. Leu
> jl...@mi...
> 
> 
>
-------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for
> FREE
> LinuxWorld Reader's Choice Award Winner for best
> database on Linux.
>
http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
>
https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
>  


		
______________________________________________
Renovamos el Correo Yahoo!: ¡100 MB GRATIS!
Nuevos servicios, más seguridad
http://correo.yahoo.es

Re: [mpls-linux-general] Router Performance

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

On Fri, Oct 29, 2004 at 10:08:47AM +0700, bla...@in... wrote:
>=20
> I want to test performance MPLS linux router compare with non MPLS linux
> router.
> does anyone ever test it??
> give some advise pls ;)

There were some results posted to this list in the past.  No performance
testing has been done on the newer 1.9xx version which have significant
changes that will affect speed.

Couple things to note:
-You will never see a speed improvement on a LER<->LSR<->LER test
 compared to normal IPv4.
-You should see speed improvemtns for Many LERs connected to one LSR.
-Waw label switching should be faster then IPv4 routing (you would need
 a traffic generator that could create labeled packets.)

>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_idU88&alloc_id=12065&op?k
> _______________________________________________
> 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] Router Performance

[<bla...@in...>]

I want to test performance MPLS linux router compare with non MPLS linux
router.
does anyone ever test it??
give some advise pls ;)

Re: [mpls-linux-general] Error when I followed "README.quagga-mpls"

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

Make sure to do a ./update-autotools before doing ./configure.

 On Wed, Oct 27, 2004 at 11:09:45PM +0800, ?????x wrote:
> I follow the steps described in "README.quagga-mpls", I apply the patch, then modify 'create-links', then execute "create-links", then I went back to the top of the quagga source, and I type
> 
> ./configure
> make clean
> 
> Then I got some error Messages:
> 
> config.status: creating bgpd/Makefile
> config.status: executing depfiles commands
> make[1]: Leaving directory `/usr/local/src/quagga-0.96.4/bgpd'
> make[1]: Entering directory `/usr/local/src/quagga-0.96.4/bgpd'
> test -z "libbgp.a" || rm -f libbgp.a
> test -z "bgpd" || rm -f bgpd
> rm -f *.o core *.core
> make[1]: Leaving directory `/usr/local/src/quagga-0.96.4/bgpd'
> Making clean in ldpd
> make[1]: Entering directory `/usr/local/src/quagga-0.96.4/ldpd'
> make[1]: *** No rule to make target `clean'.  Stop.
> make[1]: Leaving directory `/usr/local/src/quagga-0.96.4/ldpd'
> make: *** [clean-recursive] Error 1
> 
> Is there anything wrong?
> 
> 
> Timothy
> 
> 2004.10.26
> 

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

Re: [mpls-linux-general] Help me to understand

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

He is pointing out a bug that removing a MPLS iptables entry will not stop
traffic from flowing on the LSP (until the route cache entry expires)

On Wed, Oct 27, 2004 at 04:46:54PM +0200, windtim@libero.i wrote:
> Hi James,
>             i'm reading mail on the mailing list and i've not understood =
your
> answer to one (Antonio Mancuso 11/8/03):=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
>  Hello,=20
>  on my test network I noticed a strange bahaviour
>  regarding MPLS traffic.
> =20
>  Here my problem:
> =20
>  Host A: 192.168.10.10   ------------>    Host B: 192.168.10.20
> =20
> =20
>  On host A:
>  - mplsadm2 -A -O 0
>    Key: 0x00000385
>    Out Segment add: Success
> =20
>  - mplsadm2 -O 0x385 -o push:gen:170:set:tap0:ipv4:192.168.10.20
>    Out Instr: Success
> =20
>  - more /proc/net/mpls_out
>    0x00000385 0/0/0 1 PUSH(gen 170) SET(tap0,192.168.10.20)
> =20
>  Now if I do, from A: ping 192.168.10.20 on B I receive normal ethernet p=
acket
>  and that"s fine.
> =20
>  - iptables -t mangle -A OUTPUT -d 192.168.10.20 -j MPLS --set-key 0x385
> =20
>  Now, as expected, on B I receive MPLS packet marked with 170 as label.
>  - more /proc/net/mpls_out
>    0x00000385 70/5880/0 2 PUSH(gen 170) SET(tap0,192.168.10.20)
> =20
>  Now If I remove the iptables entry in this way:
>  - iptables -t mangle -D OUTPUT -d 192.168.10.20 -j MPLS --set-key 0x385
> =20
>  on host B I still receive MPLS packet marked with 170 as label
>  and
> =20
>  - more /proc/net/mpls_out
>    0x00000385 183/15372/0 2 PUSH(gen 170) SET(tap0,192.168.10.20)
> =20
>  the counter continue to go up.
> =20
> What's the strange thing? He first time set host A to give packets to hos=
t B
> with MPLS label 0x7, after remove this setting but there is outgoing labe=
l space
> setted yet.
> Can you help me to understand?
>=20
>=20
>=20
> ____________________________________________________________
> Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazio=
ne.=20
> Abbonati subito su http://www.libero.it=20
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by:
> Sybase ASE Linux Express Edition - download now for FREE
> LinuxWorld Reader's Choice Award Winner for best database on Linux.
> http://ads.osdn.com/?ad_idU88&alloc_id=12065&op?k
> _______________________________________________
> 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] Error when I followed "README.quagga-mpls"

[<m9...@cn...>]

I follow the steps described in "README.quagga-mpls", I apply the patch, =
then modify 'create-links', then execute "create-links", then I went =
back to the top of the quagga source, and I type

./configure
make clean

Then I got some error Messages:

config.status: creating bgpd/Makefile
config.status: executing depfiles commands
make[1]: Leaving directory `/usr/local/src/quagga-0.96.4/bgpd'
make[1]: Entering directory `/usr/local/src/quagga-0.96.4/bgpd'
test -z "libbgp.a" || rm -f libbgp.a
test -z "bgpd" || rm -f bgpd
rm -f *.o core *.core
make[1]: Leaving directory `/usr/local/src/quagga-0.96.4/bgpd'
Making clean in ldpd
make[1]: Entering directory `/usr/local/src/quagga-0.96.4/ldpd'
make[1]: *** No rule to make target `clean'.  Stop.
make[1]: Leaving directory `/usr/local/src/quagga-0.96.4/ldpd'
make: *** [clean-recursive] Error 1

Is there anything wrong?


Timothy

2004.10.26

[mpls-linux-general] iptables and label space

[windtim\@libero\.i <wi...@li...>]

Hi James,
           i'm reading ma/iling list about MPLS for Linux and =
i've found a mail
that i've not understood  Antonio Mancuso 11/
=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
 Hell=
o,
 on my test network I noticed a strange bahaviour
 regarding MPLS tr=
affic.
 
 Here my problem:
 
 Host A: 192.168.10.10   ------------>  =
  Host B: 192.168.10.20
 
 
 On host A:
 - mplsadm2 -A -O 0
   Key: =
0x00000385
   Out Segment add: Success
 
 - mplsadm2 -O 0x385 -o push:=
gen:170:set:tap0:ipv4:192.168.10.20
   Out Instr: Success
 
 - more /p=
roc/net/mpls_out
   0x00000385 0/0/0 1 PUSH(gen 170) SET(tap0,192.168.10=
.20)
 
 Now if I do, from A: ping 192.168.10.20 on B I receive normal e=
thernet packet
 and that"s fine.
 
 - iptables -t mangle -A OUTPUT -d =
192.168.10.20 -j MPLS --set-key 0x385
 
 Now, as expected, on B I recei=
ve MPLS packet marked with 170 as label.
 - more /proc/net/mpls_out
   =
0x00000385 70/5880/0 2 PUSH(gen 170) SET(tap0,192.168.10.20)
 
 Now If =
I remove the iptables entry in this way:
 - iptables -t mangle -D OUTPUT=
 -d 192.168.10.20 -j MPLS --set-key 0x385
 
 on host B I still receive =
MPLS packet marked with 170 as label
 and
 
 - more /proc/net/mpls_out=

   0x00000385 183/15372/0 2 PUSH(gen 170) SET(tap0,192.168.10.20)
 =0D
=
 the counter continue to go up.
 
 
 So, do you think this is a right =
behaviour?
 If yes, how can I stop host A to sending MPLS packets withou=
t removing the
 mpls key in /proc/net/mpls_out, but just acting on iptab=
les?

Which is the problem? Also if he removes the rules there is outgo=
ing lable space
with label 170 setted yet. For me it is normal that host=
 B receives packets with
MPLS header, isn't it? Could you explain me why=
?
Thanks in advance for the help=0A=0A=0A=0A____________________________=
________________________________=0ALibero ADSL: navighi gratis a 1.2 Mega=
, senza canone e costi di attivazione. =0AAbbonati subito su http://www.l=
ibero.it =0A