From danielcamara at gmail.com  Wed Apr  6 07:34:12 2011
From: danielcamara at gmail.com (Daniel Camara)
Date: Wed, 6 Apr 2011 16:34:12 +0200
Subject: [Xorp-users] installation problem xorp 1.8.3
Message-ID: <BANLkTik1ayCdnXs+KwdAWPBKW1VYuTgJrg@mail.gmail.com>

Hi people,

I am receiving the following message when I call "scons install"  from
/usr/local/xorp in a linux ubuntu (kernel  2.6.20 and xorp 1.8.3) .

Install file: "obj/i686-pc-linux-gnu/cli/tools/cli_generic" as
"obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic"
Chmod("lib/xorp/bin/cli_generic", 0755)
scons: *** [obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic]
lib/xorp/bin/cli_generic: No such file or directory
scons: building terminated because of errors.
root at oscar:/usr/local/xorp# pwd
/usr/local/xorp

I installed all the packets and everything else. Some one has any idea? The
directory
"/usr/local/xorp/obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic"
exists but not the /usr/local/xorp/lib (or  /lib/xorp I don't know the
structure :( )

 Any ideas are welcome :).

     Best regards...

              Daniel Camara
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20110406/b3cc02ec/attachment.html 

From jcoco at meccorp.mec.edu  Wed Apr  6 07:42:12 2011
From: jcoco at meccorp.mec.edu (Joe Coco)
Date: Wed, 6 Apr 2011 10:42:12 -0400
Subject: [Xorp-users] installation problem xorp 1.8.3
In-Reply-To: <BANLkTik1ayCdnXs+KwdAWPBKW1VYuTgJrg@mail.gmail.com>
References: <BANLkTik1ayCdnXs+KwdAWPBKW1VYuTgJrg@mail.gmail.com>
Message-ID: <65A6944974518C4B9C0E818727B559E94D07F929@mecexchange2007.meccorp.mec.edu>

I've always ran 'scons install' from the source directory...

-- Joe

From: xorp-users-bounces at xorp.org [mailto:xorp-users-bounces at xorp.org] On Behalf Of Daniel Camara
Sent: Wednesday, April 06, 2011 10:34 AM
To: xorp-users at xorp.org
Subject: [Xorp-users] installation problem xorp 1.8.3

Hi people,

I am receiving the following message when I call "scons install"  from /usr/local/xorp in a linux ubuntu (kernel  2.6.20 and xorp 1.8.3) .

Install file: "obj/i686-pc-linux-gnu/cli/tools/cli_generic" as "obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic"
Chmod("lib/xorp/bin/cli_generic", 0755)
scons: *** [obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic] lib/xorp/bin/cli_generic: No such file or directory
scons: building terminated because of errors.
root at oscar:/usr/local/xorp# pwd
/usr/local/xorp

I installed all the packets and everything else. Some one has any idea? The directory "/usr/local/xorp/obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic"  exists but not the /usr/local/xorp/lib (or  /lib/xorp I don't know the structure :( )

 Any ideas are welcome :).

     Best regards...

              Daniel Camara
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20110406/a5b9c5eb/attachment.html 

From danielcamara at gmail.com  Wed Apr  6 07:54:06 2011
From: danielcamara at gmail.com (Daniel Camara)
Date: Wed, 6 Apr 2011 16:54:06 +0200
Subject: [Xorp-users] installation problem xorp 1.8.3
In-Reply-To: <65A6944974518C4B9C0E818727B559E94D07F929@mecexchange2007.meccorp.mec.edu>
References: <BANLkTik1ayCdnXs+KwdAWPBKW1VYuTgJrg@mail.gmail.com>
	<65A6944974518C4B9C0E818727B559E94D07F929@mecexchange2007.meccorp.mec.edu>
Message-ID: <BANLkTi=QZxw4At-1NuHgzFn-a5VS8AUafA@mail.gmail.com>

THANKS Joe :)

 I am stupid, at some point I understood I should put everyting in the
/lib/xorp, I am stupid.  From the source, when I call the  scons install it
works :).

 Best regards...

     Daniel

On Wed, Apr 6, 2011 at 4:42 PM, Joe Coco <jcoco at meccorp.mec.edu> wrote:

>  I've always ran 'scons install' from the source directory...
>
>
>
> -- Joe
>
>
>
> *From:* xorp-users-bounces at xorp.org [mailto:xorp-users-bounces at xorp.org] *On
> Behalf Of *Daniel Camara
> *Sent:* Wednesday, April 06, 2011 10:34 AM
> *To:* xorp-users at xorp.org
> *Subject:* [Xorp-users] installation problem xorp 1.8.3
>
>
>
> Hi people,
>
> I am receiving the following message when I call "scons install"  from
> /usr/local/xorp in a linux ubuntu (kernel  2.6.20 and xorp 1.8.3) .
>
> Install file: "obj/i686-pc-linux-gnu/cli/tools/cli_generic" as
> "obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic"
> Chmod("lib/xorp/bin/cli_generic", 0755)
> scons: *** [obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic]
> lib/xorp/bin/cli_generic: No such file or directory
> scons: building terminated because of errors.
> root at oscar:/usr/local/xorp# pwd
> /usr/local/xorp
>
> I installed all the packets and everything else. Some one has any idea? The
> directory
> "/usr/local/xorp/obj/i686-pc-linux-gnu/cli/tools/lib/xorp/bin/cli_generic"
> exists but not the /usr/local/xorp/lib (or  /lib/xorp I don't know the
> structure :( )
>
>  Any ideas are welcome :).
>
>      Best regards...
>
>               Daniel Camara
>



-- 
    Best regards...

                     Daniel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20110406/4f8d5aeb/attachment.html 

From Florian.Thiessenhusen at admeritia.de  Tue Apr 26 01:17:35 2011
From: Florian.Thiessenhusen at admeritia.de (Florian Thiessenhusen)
Date: Tue, 26 Apr 2011 10:17:35 +0200
Subject: [Xorp-users] QoS on XORP
Message-ID: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA31@srvmail01.admeritia.lan>

Hi,

i found nothing concerning configuring QoS on the XORP-Website.

Is QoS supported on XORP? An if yes, how I have to configure it?

Florian

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4598 bytes
Desc: S/MIME Cryptographic Signature
Url : http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20110426/10247681/attachment.bin 

From jmitchell at ll.mit.edu  Tue Apr 26 07:04:15 2011
From: jmitchell at ll.mit.edu (Jeff Mitchell)
Date: Tue, 26 Apr 2011 10:04:15 -0400
Subject: [Xorp-users] QoS on XORP
In-Reply-To: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA31@srvmail01.admeritia.lan>
References: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA31@srvmail01.admeritia.lan>
Message-ID: <4DB6D0DF.9000901@ll.mit.edu>

On 04/26/2011 04:17 AM, Florian Thiessenhusen wrote:
> Hi,
>
> i found nothing concerning configuring QoS on the XORP-Website.
>
> Is QoS supported on XORP? An if yes, how I have to configure it?

AFAIK there isn't any explicit QoS support in XORP -- you need to 
configure this using your relevant OS's QoS capabilities.

(As a side note, my guess is that there generally can't be QoS support 
in XORP -- XORP manages the kernel routing tables and speaks some 
protocols but the queuing is all performed by the kernel.)

--Jeff


From Florian.Thiessenhusen at admeritia.de  Tue Apr 26 07:26:33 2011
From: Florian.Thiessenhusen at admeritia.de (Florian Thiessenhusen)
Date: Tue, 26 Apr 2011 16:26:33 +0200
Subject: [Xorp-users] QoS on XORP
In-Reply-To: <4DB6D0DF.9000901@ll.mit.edu>
References: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA31@srvmail01.admeritia.lan>
	<4DB6D0DF.9000901@ll.mit.edu>
Message-ID: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA97@srvmail01.admeritia.lan>

Hi Jeff,

thanks fort he response.

I was confused, because the XORP Website shows the following:
"XORP's CLI can be extended to encompass additional router functionality such as queue manangement, _QoS configuration_, firewalls, NATs and DHCP configuration."


Florian


-----Urspr?ngliche Nachricht-----
Von: Jeff Mitchell [mailto:jmitchell at ll.mit.edu] 
Gesendet: Dienstag, 26. April 2011 16:04
An: Florian Thiessenhusen
Cc: xorp-users at xorp.org
Betreff: Re: [Xorp-users] QoS on XORP

On 04/26/2011 04:17 AM, Florian Thiessenhusen wrote:
> Hi,
>
> i found nothing concerning configuring QoS on the XORP-Website.
>
> Is QoS supported on XORP? An if yes, how I have to configure it?

AFAIK there isn't any explicit QoS support in XORP -- you need to 
configure this using your relevant OS's QoS capabilities.

(As a side note, my guess is that there generally can't be QoS support 
in XORP -- XORP manages the kernel routing tables and speaks some 
protocols but the queuing is all performed by the kernel.)

--Jeff
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4598 bytes
Desc: S/MIME Cryptographic Signature
Url : http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20110426/922e1083/attachment.bin 

From jmitchell at ll.mit.edu  Tue Apr 26 07:47:12 2011
From: jmitchell at ll.mit.edu (Jeff Mitchell)
Date: Tue, 26 Apr 2011 10:47:12 -0400
Subject: [Xorp-users] QoS on XORP
In-Reply-To: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA97@srvmail01.admeritia.lan>
References: <D79FDDD8143E7B4CBCDB375A698BDA07150C91EA31@srvmail01.admeritia.lan>
	<4DB6D0DF.9000901@ll.mit.edu>
	<D79FDDD8143E7B4CBCDB375A698BDA07150C91EA97@srvmail01.admeritia.lan>
Message-ID: <4DB6DAF0.5060907@ll.mit.edu>

On 04/26/2011 10:26 AM, Florian Thiessenhusen wrote:
> Hi Jeff,
>
> thanks fort he response.
>
> I was confused, because the XORP Website shows the following:
> "XORP's CLI can be extended to encompass additional router functionality such as queue manangement, _QoS configuration_, firewalls, NATs and DHCP configuration."

In theory the CLI can be extended to do so.

I'm not aware of anyone having actually done that.

--Jeff


From greearb at candelatech.com  Thu Apr 28 09:35:39 2011
From: greearb at candelatech.com (Ben Greear)
Date: Thu, 28 Apr 2011 09:35:39 -0700
Subject: [Xorp-users] multicast: Fix critical bug with mcast routing tables
	in Linux.
Message-ID: <4DB9975B.5010806@candelatech.com>


If anyone is trying to use xorp 1.8.3 with multicast routing on Linux,
and using kernel 2.6.26 or higher, it will not work without the
patch I just committed.  Please use the top-of-tree or manually
apply this patch to your xorp source.


-------- Original Message --------
Subject: [Xorp-hackers] [greearb/xorp.ct] 4282c5: multicast: Fix critical	bug with mcast routing ta...
Date: Thu, 28 Apr 2011 09:31:37 -0700
From: noreply at github.com
To: xorp-hackers at xorp.org

Branch: refs/heads/master
Home:   https://github.com/greearb/xorp.ct

Commit: 4282c51b9fdf954280fdfb25d49d0a9c4a7fa12d
     https://github.com/greearb/xorp.ct/commit/4282c51b9fdf954280fdfb25d49d0a9c4a7fa12d
Author: Ben Greear <greearb at candelatech.com>
Date:   2011-04-28 (Thu, 28 Apr 2011)

Changed paths:
   M xorp/RELEASE_NOTES
   M xorp/fea/mfea_mrouter.cc

Log Message:
-----------
multicast:  Fix critical bug with mcast routing tables in Linux.

When adding the virtual routing tables for multicast, I messed up
the default routing table for more normal users.  It was set to
254 instead of 253.  This broke multicast routing for users on
recent (2.6.36 and later) kernels.  We did not catch this in
our testing because we always specify the routing table.

Reported-by: Jeff Mitchell <jmitchell at ll.mit.edu>
Tested-by: Jeff Mitchell <jmitchell at ll.mit.edu>
Signed-off-by: Ben Greear <greearb at candelatech.com>


Patch:


diff --git a/xorp/fea/mfea_mrouter.cc b/xorp/fea/mfea_mrouter.cc
index 2d74f1a..e554c12 100644
--- a/xorp/fea/mfea_mrouter.cc
+++ b/xorp/fea/mfea_mrouter.cc
@@ -101,14 +101,10 @@ bool new_mcast_tables_api = false;
   * here. --Ben
   */

-/* NOTE:  This needs to go away as soon as we get the official support
- * merged into the linux kernel.
- */
-
  // Assume supported until we know otherwise.
  bool supports_mcast_tables = true;

-#define DFLT_MROUTE_TBL 254  /* 'main' routing table id in Linux */
+#define DFLT_MROUTE_TBL 253  /* 'default' routing table id in Linux */


  // Support for multiple routing tables.


Thanks,
Ben

_______________________________________________
Xorp-hackers mailing list
Xorp-hackers at icir.org
http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-hackers


From kiran.nagaraja at rutgers.edu  Thu Apr 28 11:10:42 2011
From: kiran.nagaraja at rutgers.edu (Kiran Nagaraja)
Date: Thu, 28 Apr 2011 14:10:42 -0400
Subject: [Xorp-users] Click as forwarding engine
Message-ID: <4DB9ADA2.4080904@rutgers.edu>

Hi all,

 From looking at the archives (and use attempt), I gather that recent 
versions of XORP no longer support Click as forwarding engine, and that 
some previous version combination of the two might work (xorp 1.52 and 
click 1.6.0 ?) for those interested.

We're looking to run BGP and a custom protocol, modules of which we're 
looking to build as Click elements. I quite like XORP framework and 
wanted to find out: 1.) if we're losing anything (major) by backtracking 
to older version, and 2.) if there's considerable interest to 
resurrecting Click under XORP. Any details on how much the two 
components have deviated in their interfaces?


Thanks in advance for any comments,
- Kiran
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20110428/40624b89/attachment.html 

From greearb at candelatech.com  Thu Apr 28 11:15:14 2011
From: greearb at candelatech.com (Ben Greear)
Date: Thu, 28 Apr 2011 11:15:14 -0700
Subject: [Xorp-users] Click as forwarding engine
In-Reply-To: <4DB9ADA2.4080904@rutgers.edu>
References: <4DB9ADA2.4080904@rutgers.edu>
Message-ID: <4DB9AEB2.6060309@candelatech.com>

On 04/28/2011 11:10 AM, Kiran Nagaraja wrote:
>   Hi all,
>
>  From looking at the archives (and use attempt), I gather that recent
> versions of XORP no longer support Click as forwarding engine, and that
> some previous version combination of the two might work (xorp 1.52 and
> click 1.6.0 ?) for those interested.
>
> We're looking to run BGP and a custom protocol, modules of which we're
> looking to build as Click elements. I quite like XORP framework and
> wanted to find out: 1.) if we're losing anything (major) by backtracking
> to older version, and 2.) if there's considerable interest to
> resurrecting Click under XORP. Any details on how much the two
> components have deviated in their interfaces?

At the least, if you use an older version, it's a lot less likely
anyone will be able to offer much help with bugs you may find.

I would be happy to accept patches to code and documentation to get click working in
the latest code.  I have no idea how much things have deviated
since click last worked, but likely to too much.

Thanks,
Ben

-- 
Ben Greear <greearb at candelatech.com>
Candela Technologies Inc  http://www.candelatech.com


From billaud at vmware.com  Thu Apr 28 20:39:03 2011
From: billaud at vmware.com (Jean-Pascal Billaud)
Date: Thu, 28 Apr 2011 20:39:03 -0700
Subject: [Xorp-users] Xorp PIM-SM over GRE issue?
Message-ID: <8B1F619C9F5F454E81D90D3C161698D7033DAA4B38@EXCH-MBX-3.vmware.com>

Hey folks,

It is my first time deploying Xorp so I might have missed something obvious... please find below my network:

VM1 (192.168.1.1)    MRouter1 (192.168.1.253)
	  |                      |
        ----------------------------------
							| (192.168.1.254)
						Unicast Router
							| (192.168.2.254)
							-------------------------------------
								|					|
						MRouter2 (192.168.2.253)		VM2 (192.168.2.1)

All the entities in there are VMs BTW even that should not matter in practice. What I would like to do is to have VM1 & VM2 doing some iperf multicast going through their respective L3 MRouter as opposed to going through the Unicast Router stitching the L3 network. I am sure this is a very typical scenario.

So I have a GRE tunnel between both MRouter to have them PIM peers with the following /etc/network/interfaces:

# Configuration for MRouter1
# The primary network interface
auto eth0
iface eth0 inet static
address 192.168.1.253
netmask 255.255.255.0
network 192.168.1.0
broadcast 192.168.1.255
gateway 192.168.1.254

auto tun0
iface tun0 inet static
address 192.168.10.1
netmask 255.255.255.0
broadcast 192.168.10.255
up ifconfig tun0 multicast
pointopoint 192.168.10.2
pre-up iptunnel add tun0 mode gre remote 192.168.2.253 local 192.168.1.253 ttl 255
post-down iptunnel del tun0

I am able to ping MRouter2 through the tunnel and I can observe the GRE frames at the Unicast Router using tcpdump. Obviously the MRouter2 interface file differs a bit due to ip address...

Then here is my Xorp configuration on MRouter1... The MRouter2's one is slightly different for the same reasons as above with the interface file:

interfaces {
           interface eth0 {
                     vif eth0 {
                         address 192.168.1.253 {
                                  prefix-length: 24
                         }
                     }
           }
           interface tun0 {
                     vif tun0 {
                         address 192.168.10.1 {
                                  prefix-length: 24
                                  multicast-capable: true
                         }
                     }
           }
}
plumbing {
         mfea4 {
               interface eth0 {
                         vif eth0 {
                         }
               }
               interface tun0 {
                         vif tun0 {
                         }
               }
               interface "register_vif" {
                         vif "register_vif" {
                         }
               }
               traceoptions {
                            flag all {
                                 disable: false
                            }
               }
         }
}

protocols {
         igmp {
              interface eth0 {
                        vif eth0 {
                            version: 3
                            query-interval: 30
                        }
              }
              interface tun0 {
                        vif tun0 {
                        }
              }
              traceoptions {
                           flag all {
                                disable: false
                           }
              }
         }

         pimsm4 {
                interface eth0 {
                          vif eth0 {
                          }
                }
                interface tun0 {
                          vif tun0 {
                              hello-period: 10
                          }
                }
                interface "register_vif" {
                          vif "register_vif" {
                          }
                }

                static-rps {
                           rp 192.168.10.1 {
                              group-prefix 224.0.0.0/4 {
                              }
                           }
                }
                traceoptions {
                             flag all {
                                  disable: false
                             }
                }
         }
         static {
                mrib-route 0.0.0.0/0 {
                           next-hop: 192.168.10.2
                }
         }
}

Now the issue is that while Xorp is running PIM-SM on MRouter1, it sends HELLO through tun0 interface but for some reasons it uses the eth0 IP address and there is no GRE done at all... Looking at tcpdump on eth0 on MRouter1 we can see non-GRE PIM frames:

20:33:45.443462 IP (tos 0xc0, ttl 1, id 18031, offset 0, flags [none], proto PIM (103), length 64)
    192.168.1.253 > 224.0.0.13: PIMv2, length 44
        Hello, cksum 0xcfa6 (correct)
          Hold Time Option (1), length 2, Value: 35s
            0x0000:  0023
          LAN Prune Delay Option (2), length 4, Value:
            T-bit=0, LAN delay 500ms, Override interval 2500ms
            0x0000:  01f4 09c4
          DR Priority Option (19), length 4, Value: 1
            0x0000:  0000 0001
          Generation ID Option (20), length 4, Value: 0x3a13fe69
            0x0000:  3a13 fe69
          Address List Option (24), length 6, Value:
            192.168.10.1
            0x0000:  0100 c0a8 0a01

>From Xorp logs:

[ 2011/04/28 20:17:20.581125 INFO xorp_igmp MLD6IGMP ] Interface enabled: Vif[tun0] pif_index: 12 vif_index: 1 addr: 192.168.1.253 subnet: 192.168.1.253/32 broadcast: 0.0.0.0 peer: 192.168.2.253 addr: 192.168.10.1 subnet: 192.168.10.0/24 broadcast: 0.0.0.0 peer: 192.168.10.1 Flags: P2P MULTICAST UNDERLYING_VIF_UP MTU: 1476 DOWN IPv4 ENABLED

[ 2011/04/28 20:17:36.304614 TRACE xorp_pimsm4 PIM ] pim_send: TX PIM_HELLO from 192.168.1.253 to 224.0.0.13 on vif tun0
[ 2011/04/28 20:17:46.306477 TRACE xorp_pimsm4 PIM ] pim_send: TX PIM_HELLO from 192.168.1.253 to 224.0.0.13 on vif tun0

root at promd-1n-dhcp198> show pim interface
Interface    State    Mode   V PIMstate Priority DRaddr          Neighbors
eth0         UP       Sparse 2 DR              1 192.168.1.253           0
tun0         UP       Sparse 2 DR              1 192.168.1.253           0
register_vif UP       Sparse 2 DR              1 192.168.1.253           0

Any help would be really appreciated.

Thanks,
--jp


From greearb at candelatech.com  Thu Apr 28 21:06:41 2011
From: greearb at candelatech.com (Ben Greear)
Date: Thu, 28 Apr 2011 21:06:41 -0700
Subject: [Xorp-users] Xorp PIM-SM over GRE issue?
In-Reply-To: <8B1F619C9F5F454E81D90D3C161698D7033DAA4B38@EXCH-MBX-3.vmware.com>
References: <8B1F619C9F5F454E81D90D3C161698D7033DAA4B38@EXCH-MBX-3.vmware.com>
Message-ID: <4DBA3951.9070404@candelatech.com>

On 04/28/2011 08:39 PM, Jean-Pascal Billaud wrote:
> Hey folks,
>
> It is my first time deploying Xorp so I might have missed something obvious... please find below my network:

Please make sure you are using top-of-tree, not 1.8.3.  I just fixed a bug earlier today
that broke multicast routing in many cases on Linux.  But, not sure if that is what is wrong in
your case or not.

Please let us know which version of xorp you are using and what OS/kernel.

Thanks,
Ben

-- 
Ben Greear <greearb at candelatech.com>
Candela Technologies Inc  http://www.candelatech.com


From billaud at vmware.com  Thu Apr 28 22:16:58 2011
From: billaud at vmware.com (Jean-Pascal Billaud)
Date: Thu, 28 Apr 2011 22:16:58 -0700
Subject: [Xorp-users] Xorp PIM-SM over GRE issue?
In-Reply-To: <4DBA3951.9070404@candelatech.com>
References: <8B1F619C9F5F454E81D90D3C161698D7033DAA4B38@EXCH-MBX-3.vmware.com>
	<4DBA3951.9070404@candelatech.com>
Message-ID: <8B1F619C9F5F454E81D90D3C161698D7033DAA4B43@EXCH-MBX-3.vmware.com>

Hey Ben,

I am running ubuntu 10.10 server 64bit 

Linux promd-1n-dhcp198 2.6.35-22-server #33-Ubuntu SMP Sun Sep 19 20:48:58 UTC 2010 x86_64 GNU/Linux

$>git log
commit 7a4f917e23ff01eeef26513404b3a4913af131da
Merge: 783b88d 323cbc6
Author: Ben Greear <greearb at candelatech.com>
Date:   Sat Apr 23 08:32:02 2011 -0700

    Merge branch 'master' of github.com:greearb/xorp.ct

commit 783b88d11cd3e8463592974cb74c77e67aabdfc3
Author: Ben Greear <greearb at candelatech.com>
Date:   Sat Apr 23 08:29:53 2011 -0700

    BSD:  Fix bug with adding new virtual interface.

    BSD has the bad habit of re-using interface ifindexes
    if you delete and then re-add an interface.  Instead of
    asserting, delete the old interface with the duplicate
    ifindex before adding the new one.

    Signed-off-by: Ben Greear <greearb at candelatech.com>

So I am already past 1.8.3... I am going to pull to get your latest patch and test again.

Thanks,
--jp

> -----Original Message-----
> From: Ben Greear [mailto:greearb at candelatech.com]
> Sent: Thursday, April 28, 2011 9:07 PM
> To: Jean-Pascal Billaud
> Cc: xorp-users at xorp.org
> Subject: Re: [Xorp-users] Xorp PIM-SM over GRE issue?
> 
> On 04/28/2011 08:39 PM, Jean-Pascal Billaud wrote:
> > Hey folks,
> >
> > It is my first time deploying Xorp so I might have missed something
> obvious... please find below my network:
> 
> Please make sure you are using top-of-tree, not 1.8.3.  I just fixed a
> bug earlier today
> that broke multicast routing in many cases on Linux.  But, not sure if
> that is what is wrong in
> your case or not.
> 
> Please let us know which version of xorp you are using and what
> OS/kernel.
> 
> Thanks,
> Ben
> 
> --
> Ben Greear <greearb at candelatech.com>
> Candela Technologies Inc  http://www.candelatech.com


From billaud at vmware.com  Fri Apr 29 10:40:24 2011
From: billaud at vmware.com (Jean-Pascal Billaud)
Date: Fri, 29 Apr 2011 10:40:24 -0700
Subject: [Xorp-users] Xorp PIM-SM over GRE issue?
In-Reply-To: <4DBAB03C.4030104@ll.mit.edu>
References: <8B1F619C9F5F454E81D90D3C161698D7033DAA4B38@EXCH-MBX-3.vmware.com>
	<4DBA3951.9070404@candelatech.com>
	<8B1F619C9F5F454E81D90D3C161698D7033DAA4B43@EXCH-MBX-3.vmware.com>
	<4DBAB03C.4030104@ll.mit.edu>
Message-ID: <8B1F619C9F5F454E81D90D3C161698D7033DAA4BF3@EXCH-MBX-3.vmware.com>

So I sync'd top of the tree and I still see the same issue. Do you guys see something wrong with my setup? Do I have the right expectations?

One weird thing is that the tunnel address 192.168.10.1 is in the Address List Option of the PIM HELLO message... 

[ 2011/04/29 10:28:22.210023 TRACE xorp_pimsm4 PIM ] pim_send: TX PIM_HELLO from 192.168.1.253 to 224.0.0.13 on vif eth0
[ 2011/04/29 10:28:23.6958  INFO xorp_rtrmgr:10153 RTRMGR rtrmgr/task.cc:2242 run_task ] No more tasks to run
[ 2011/04/29 10:28:23.434028 TRACE xorp_pimsm4 PIM ] pim_send: TX PIM_HELLO from 192.168.1.253 to 224.0.0.13 on vif tun0 <=== why not from 192.168.10.1?

20:33:45.443462 IP (tos 0xc0, ttl 1, id 18031, offset 0, flags [none], proto PIM (103), length 64)
    192.168.1.253 > 224.0.0.13: PIMv2, length 44
        Hello, cksum 0xcfa6 (correct)
          Hold Time Option (1), length 2, Value: 35s
            0x0000:  0023
          LAN Prune Delay Option (2), length 4, Value:
            T-bit=0, LAN delay 500ms, Override interval 2500ms
            0x0000:  01f4 09c4
          DR Priority Option (19), length 4, Value: 1
            0x0000:  0000 0001
          Generation ID Option (20), length 4, Value: 0x3a13fe69
            0x0000:  3a13 fe69
          Address List Option (24), length 6, Value:
            192.168.10.1
            0x0000:  0100 c0a8 0a01

Note that tun0 (192.168.10.1) is built on top of eth0 (192.168.1.253)...

Also just to confirm the GRE tunnel is well in place, below is a tcpdump on eth0:

10:24:39.617109 IP (tos 0x0, ttl 255, id 0, offset 0, flags [DF], proto GRE (47), length 108)
    192.168.1.253 > 192.168.2.253: GREv0, Flags [none], length 88
        IP (tos 0x0, ttl 64, id 0, offset 0, flags [DF], proto ICMP (1), length 84)
    192.168.10.1 > 192.168.10.2: ICMP echo request, id 10034, seq 18, length 64

--jp

> -----Original Message-----
> From: Jeff Mitchell [mailto:jmitchell at ll.mit.edu]
> Sent: Friday, April 29, 2011 5:34 AM
> To: Jean-Pascal Billaud
> Subject: Re: [Xorp-users] Xorp PIM-SM over GRE issue?
> 
> On 4/29/2011 1:16 AM, Jean-Pascal Billaud wrote:
> > Hey Ben,
> >
> > I am running ubuntu 10.10 server 64bit
> >
> > Linux promd-1n-dhcp198 2.6.35-22-server #33-Ubuntu SMP Sun Sep 19
> 20:48:58 UTC 2010 x86_64 GNU/Linux
> >
> > $>git log
> > commit 7a4f917e23ff01eeef26513404b3a4913af131da
> > Merge: 783b88d 323cbc6
> > Author: Ben Greear<greearb at candelatech.com>
> > Date:   Sat Apr 23 08:32:02 2011 -0700
> 
> You need to update git again. Ubuntu 10.04 LTS is not affected but
> 10.10
> may very well be. Update to 4282c51b or newer.
> 
> --Jeff