From ramu.avula at gmail.com Sat Jun 3 08:47:52 2006 From: ramu.avula at gmail.com (Ramu k) Date: Sat, 3 Jun 2006 21:17:52 +0530 Subject: [Xorp-users] VLAN Support in XORP Message-ID: Hi, I want to use VLANS with Xorp. Does XORP supports tag based VLANS. if it supports how can configure XORP for VLANS. I am working on multicast applicatio and i have created 20 VLANS with different IP Addreses in client machine. I am able generate 20 IGMP V3 requests for the same group but not able get responses from XORP router. MY lab setup is like this. MCASTSERVER client with 20 VLANS. A-------------------XORP-------------B Can anybody help me in this regard. Thanks In Advance. Ramuk -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060603/046693fc/attachment.html From pavlin at icir.org Sat Jun 3 09:49:21 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Sat, 03 Jun 2006 09:49:21 -0700 Subject: [Xorp-users] VLAN Support in XORP In-Reply-To: Message from "Ramu k" of "Sat, 03 Jun 2006 21:17:52 +0530." Message-ID: <200606031649.k53GnLeX034098@possum.icir.org> > I want to use VLANS with Xorp. Does XORP supports tag based VLANS. The VLAN management support is not integrated in XORP (yet). Currently, you would have to configure/manage the VLANs outside of XORP. Once you have the VLANs configured by using the tools provided by the underlying system, you can use/configure the vlan interfaces in XORP as any other interfaces. > if it supports how can configure XORP for VLANS. I am working on multicast > applicatio and i have created 20 VLANS with different IP Addreses in client > machine. > I am able generate 20 IGMP V3 requests for the same group but not able get > responses from XORP router. I should let you know that XORP doesn't support IGMPv3, but currently we are working on it. Hopefully, it should be ready for the next release (by the end of the month or so). Pavlin > MY lab setup is like this. > MCASTSERVER client with 20 VLANS. > A-------------------XORP-------------B > > Can anybody help me in this regard. > > Thanks In Advance. > > Ramuk > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From allan at vyatta.com Sat Jun 3 12:17:29 2006 From: allan at vyatta.com (Allan Leinwand) Date: Sat, 3 Jun 2006 12:17:29 -0700 (PDT) Subject: [Xorp-users] VLAN Support in XORP Message-ID: <637496.211149362249934.JavaMail.root@mail.vyatta.com> Hi Ramu, We've taken the step to integrate VLANs into the Vyatta OFR (www.vyatta.com/download), but as Pavlin points out we also don't have IGMPv3 support as yet. If you'd like to try our image and then give us feedback on the VLAN support, we'd welcome the input. Take care, Allan ----- Original Message ----- From: Ramu k To: Xorp-users at xorp.org Sent: Saturday, June 3, 2006 8:47:52 AM GMT-0800 Subject: [Xorp-users] VLAN Support in XORP Hi, I want to use VLANS with Xorp. Does XORP supports tag based VLANS. if it supports how can configure XORP for VLANS. I am working on multicast applicatio and i have created 20 VLANS with different IP Addreses in client machine. I am able generate 20 IGMP V3 requests for the same group but not able get responses from XORP router. MY lab setup is like this. MCASTSERVER client with 20 VLANS. A-------------------XORP-------------B Can anybody help me in this regard. Thanks In Advance. Ramuk From ramu.avula at gmail.com Mon Jun 5 01:23:55 2006 From: ramu.avula at gmail.com (Ramu k) Date: Mon, 5 Jun 2006 13:53:55 +0530 Subject: [Xorp-users] VLAN Support in XORP In-Reply-To: <637496.211149362249934.JavaMail.root@mail.vyatta.com> References: <637496.211149362249934.JavaMail.root@mail.vyatta.com> Message-ID: Hi Allan, I configured VLANS in client and router and both of them have FC4. Now the problem is , the client is always generating IGMPv3 report when i am trying to join a group and data is not forwarded . my /proc/net/igmp values are like this . dx Device : Count Querier Group Users Timer Reporter 1 lo : 0 V3 010000E0 1 0:00000000 0 2 eth0 : 4 V2 FB0000E0 1 0:00000000 0 010000E0 1 0:00000000 0 3 eth0.12 : 3 V3 010000E0 1 0:00000000 0 010000E0 1 0:00000000 0 how can i make eth0.12 as IGMPv3. Thanks, Ramu.k ---------- Forwarded message ---------- From: Allan Leinwand Date: Jun 4, 2006 12:47 AM Subject: Re: [Xorp-users] VLAN Support in XORP To: Ramu k Cc: Xorp-users at xorp.org Hi Ramu, We've taken the step to integrate VLANs into the Vyatta OFR ( www.vyatta.com/download), but as Pavlin points out we also don't have IGMPv3 support as yet. If you'd like to try our image and then give us feedback on the VLAN support, we'd welcome the input. Take care, Allan ----- Original Message ----- From: Ramu k To: Xorp-users at xorp.org Sent: Saturday, June 3, 2006 8:47:52 AM GMT-0800 Subject: [Xorp-users] VLAN Support in XORP Hi, I want to use VLANS with Xorp. Does XORP supports tag based VLANS. if it supports how can configure XORP for VLANS. I am working on multicast applicatio and i have created 20 VLANS with different IP Addreses in client machine. I am able generate 20 IGMP V3 requests for the same group but not able get responses from XORP router. MY lab setup is like this. MCASTSERVER client with 20 VLANS. A-------------------XORP-------------B Can anybody help me in this regard. Thanks In Advance. Ramuk -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060605/0461d705/attachment.html From allan at vyatta.com Mon Jun 5 07:12:27 2006 From: allan at vyatta.com (Allan Leinwand) Date: Mon, 5 Jun 2006 07:12:27 -0700 (PDT) Subject: [Xorp-users] VLAN Support in XORP Message-ID: <18128858.01149516747449.JavaMail.root@mail.vyatta.com> Hi Ramu, Is this using XORP 1.2 or the Vyatta OFR? As I think I mentioned earlir, neither of these images has IGMPv3 support. The OFR does have native VLAN support and that may be helpful for your implementation. Take care, Allan ----- Original Message ----- From: Ramu k To: Xorp-users at xorp.org, allan at vyatta.com Sent: Monday, June 5, 2006 1:23:55 AM GMT-0800 Subject: [Xorp-users] VLAN Support in XORP Hi Allan, I configured VLANS in client and router and both of them have FC4. Now the problem is , the client is always generating IGMPv3 report when i am trying to join a group and data is not forwarded . my /proc/net/igmp values are like this . dx Device : Count Querier Group Users Timer Reporter 1 lo : 0 V3 010000E0 1 0:00000000 0 2 eth0 : 4 V2 FB0000E0 1 0:00000000 0 010000E0 1 0:00000000 0 3 eth0.12 : 3 V3 010000E0 1 0:00000000 0 010000E0 1 0:00000000 0 how can i make eth0.12 as IGMPv3. Thanks, Ramu.k ---------- Forwarded message ---------- From: Allan Leinwand < allan at vyatta.com > Date: Jun 4, 2006 12:47 AM Subject: Re: [Xorp-users] VLAN Support in XORP To: Ramu k < ramu.avula at gmail.com > Cc: Xorp-users at xorp.org Hi Ramu, We've taken the step to integrate VLANs into the Vyatta OFR ( www.vyatta.com/download ), but as Pavlin points out we also don't have IGMPv3 support as yet. If you'd like to try our image and then give us feedback on the VLAN support, we'd welcome the input. Take care, Allan ----- Original Message ----- From: Ramu k < ramu.avula at gmail.com > To: Xorp-users at xorp.org Sent: Saturday, June 3, 2006 8:47:52 AM GMT-0800 Subject: [Xorp-users] VLAN Support in XORP Hi, I want to use VLANS with Xorp. Does XORP supports tag based VLANS. if it supports how can configure XORP for VLANS. I am working on multicast applicatio and i have created 20 VLANS with different IP Addreses in client machine. I am able generate 20 IGMP V3 requests for the same group but not able get responses from XORP router. MY lab setup is like this. MCASTSERVER client with 20 VLANS. A-------------------XORP-------------B Can anybody help me in this regard. Thanks In Advance. Ramuk From binary.chen at gmail.com Mon Jun 5 08:28:52 2006 From: binary.chen at gmail.com (Bin Chen) Date: Mon, 5 Jun 2006 23:28:52 +0800 Subject: [Xorp-users] newbie's question Message-ID: <5800c1cc0606050828u39aef25s1b3fe60c34735a82@mail.gmail.com> Hi, I wrote a test program for XORP that can accept some arguments, but after the xorp_rtrmgr starts, I edit with the field be using: create tests route4 { dport 1 } then commit, but fails with: WARNING xorp_rtrmgr:24585 XrlRtrmgrTarget +472 rtrmgr_base.cc handle_rtrmgr_0_1_apply_config_change ] Handling method for rtrmgr/0.1/apply_config_change failed: XrlCmdError 102 Command failed PARSE ERROR [line 2]: No template found in template map; Last symbol parsed was "0" My template file is : tests { targetname: txt = "test"; disable: toggle = false; id: u32; route4 @: i32{ prot: u32; } } tests { %modinfo: provides test; %modinfo: path "test/test_generated"; %modinfo: default_targetname "test"; targetname: txt = "test"; disable: toggle = false; id { %help: short "my id"; %set: xrl "$(tests.targetname)/test/1.0/set_test_id?id:u32=$(@)"; } route4 @: i32{ %help: short "Configure an IPv4 static route"; %activate: xrl "$(tests.targetname)/test/1.0/add_route4?prot:u32=(@.prot)"; %mandatory: $(@.prot); prot { %help: short "Configure the next-hop router"; %set:; } } } Why? Thanks in advance! B.C From mike at lrlart.com Mon Jun 5 08:50:10 2006 From: mike at lrlart.com (mike at lrlart.com) Date: Mon, 05 Jun 2006 08:50:10 -0700 Subject: [Xorp-users] newbie's question Message-ID: <20060605085010.8c0fa638fcef9dd13f5a2f992b1941af.bb8aca02d0.wbe@email.secureserver.net> Hi Bin, The template that you provided below doesn't have a "dport" variable defined, which would mean that rtrmgr wouldn't accept your configuration as a valid argument. However, it looks like this isn't the problem below--did you create/implement a new tests module to receive the XRL defined in your template? When you fire up the rtrmgr and the xorpsh do you see the new test commands when you type "?" in configuration mode? Also, it is useful to run the rtrmgr with the -v flag (if you aren't doing that already)--that will give you more information on what the rtrmgr is doing. Good luck--hope this helps. Mike Larson vyatta > -------- Original Message -------- > Subject: [Xorp-users] newbie's question > From: "Bin Chen" > Date: Mon, June 05, 2006 8:28 am > To: Xorp-users at xorp.org > > Hi, > > I wrote a test program for XORP that can accept some arguments, but > after the xorp_rtrmgr starts, I edit with the field be using: > > create tests route4 { > dport 1 > } > > then commit, but fails with: > > WARNING xorp_rtrmgr:24585 XrlRtrmgrTarget +472 rtrmgr_base.cc > handle_rtrmgr_0_1_apply_config_change ] Handling method for > rtrmgr/0.1/apply_config_change failed: XrlCmdError 102 Command failed > PARSE ERROR [line 2]: No template found in template map; Last symbol > parsed was "0" > > My template file is : > > tests { > targetname: txt = "test"; > disable: toggle = false; > id: u32; > route4 @: i32{ > prot: u32; > } > } > > tests { > %modinfo: provides test; > %modinfo: path "test/test_generated"; > %modinfo: default_targetname "test"; > > targetname: txt = "test"; > disable: toggle = false; > > id { > %help: short "my id"; > %set: xrl "$(tests.targetname)/test/1.0/set_test_id?id:u32=$(@)"; > } > > route4 @: i32{ > %help: short "Configure an IPv4 static route"; > > > %activate: xrl "$(tests.targetname)/test/1.0/add_route4?prot:u32=(@.prot)"; > > %mandatory: $(@.prot); > > > prot { > %help: short "Configure the next-hop router"; > %set:; > } > } > } > > Why? Thanks in advance! > > B.C > > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From pavlin at icir.org Mon Jun 5 10:01:03 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Mon, 05 Jun 2006 10:01:03 -0700 Subject: [Xorp-users] VLAN Support in XORP In-Reply-To: Message from "Ramu k" of "Mon, 05 Jun 2006 13:53:55 +0530." Message-ID: <200606051701.k55H139b056026@possum.icir.org> > I configured VLANS in client and router and both of them have FC4. > Now the problem is , the client is always generating IGMPv3 report > when i am trying to join a group and data is not forwarded . > > my /proc/net/igmp values are like this . > > dx Device : Count Querier Group Users Timer Reporter > 1 lo : 0 V3 > 010000E0 1 0:00000000 0 > 2 eth0 : 4 V2 > FB0000E0 1 0:00000000 0 > 010000E0 1 0:00000000 0 > 3 eth0.12 : 3 V3 > 010000E0 1 0:00000000 0 > 010000E0 1 0:00000000 0 > > how can i make eth0.12 as IGMPv3. If you start a multicast router connected to eth0.12, and if that router originates IGMPv2 Query messages, then eth0.12 will automatically switch to using IGMPv2. Hence, make sure that the IGMPv2 Query messages from the XORP router are reaching your client running on eth0.12. Tcpdump is your friend here. Pavlin From pavlin at icir.org Mon Jun 5 10:27:29 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Mon, 05 Jun 2006 10:27:29 -0700 Subject: [Xorp-users] newbie's question In-Reply-To: Message from mike@lrlart.com of "Mon, 05 Jun 2006 08:50:10 PDT." <20060605085010.8c0fa638fcef9dd13f5a2f992b1941af.bb8aca02d0.wbe@email.secureserver.net> Message-ID: <200606051727.k55HRTX9056295@possum.icir.org> Bin, In addition to the dport issue, you also must specify an integer value after route4. E.g.: create tests route4 10 { port 1 } Pavlin > The template that you provided below doesn't have a "dport" variable > defined, which would mean that rtrmgr wouldn't accept your > configuration as a valid argument. However, it looks like this isn't > the problem below--did you create/implement a new tests module to > receive the XRL defined in your template? When you fire up the rtrmgr > and the xorpsh do you see the new test commands when you type "?" in > configuration mode? Also, it is useful to run the rtrmgr with the -v > flag (if you aren't doing that already)--that will give you more > information on what the rtrmgr is doing. Good luck--hope this helps. > > Mike Larson > vyatta > > > -------- Original Message -------- > > Subject: [Xorp-users] newbie's question > > From: "Bin Chen" > > Date: Mon, June 05, 2006 8:28 am > > To: Xorp-users at xorp.org > > > > Hi, > > > > I wrote a test program for XORP that can accept some arguments, but > > after the xorp_rtrmgr starts, I edit with the field be using: > > > > create tests route4 { > > dport 1 > > } > > > > then commit, but fails with: > > > > WARNING xorp_rtrmgr:24585 XrlRtrmgrTarget +472 rtrmgr_base.cc > > handle_rtrmgr_0_1_apply_config_change ] Handling method for > > rtrmgr/0.1/apply_config_change failed: XrlCmdError 102 Command failed > > PARSE ERROR [line 2]: No template found in template map; Last symbol > > parsed was "0" > > > > My template file is : > > > > tests { > > targetname: txt = "test"; > > disable: toggle = false; > > id: u32; > > route4 @: i32{ > > prot: u32; > > } > > } > > > > tests { > > %modinfo: provides test; > > %modinfo: path "test/test_generated"; > > %modinfo: default_targetname "test"; > > > > targetname: txt = "test"; > > disable: toggle = false; > > > > id { > > %help: short "my id"; > > %set: xrl "$(tests.targetname)/test/1.0/set_test_id?id:u32=$(@)"; > > } > > > > route4 @: i32{ > > %help: short "Configure an IPv4 static route"; > > > > > > %activate: xrl "$(tests.targetname)/test/1.0/add_route4?prot:u32=(@.prot)"; > > > > %mandatory: $(@.prot); > > > > > > prot { > > %help: short "Configure the next-hop router"; > > %set:; > > } > > } > > } > > > > Why? Thanks in advance! > > > > B.C > > > > _______________________________________________ > > Xorp-users mailing list > > Xorp-users at xorp.org > > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users > > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From ramu.avula at gmail.com Tue Jun 6 09:02:32 2006 From: ramu.avula at gmail.com (Ramu k) Date: Tue, 6 Jun 2006 21:32:32 +0530 Subject: [Xorp-users] VLAN Support in XORP In-Reply-To: <200606051701.k55H139b056026@possum.icir.org> References: <200606051701.k55H139b056026@possum.icir.org> Message-ID: Hi Pavlin, (I am working on IPTV testing. I am experimenting XORP for calculating join latency and leave latency. If i run one multicast user join latency is around (0.3msto 0.6ms) which quite OK . but leave latency i around 3secs which is very large value.is it beacause of XORP or its a problem with my machine(FC4).) I am facing a problem in the following scenario(VLANS). My Lab setup is like below 192.168.5.7 (vlan eth0.12) A----------------------XORP-----------------------------B eth1 eth0 eth0 192.168.5.1 192.168.1.157 eth0.12 ip address is 10.0.3.3 eth0.45 ip address is 10.0.3.15 In Router machine(XORP) i configured 2 VLANS using command(eth0.12 ð0.45) VCONFIG add eth0 12 . i added this VLAN to boot.config. I am running mcast server on machine A(it has a VLAN 12 and ip 10.0.3.4). i am getting packets to eth1 of XORP. but it is not forwarding eth0.12 (ttl > 1 and added route in iptables also) Do i need to have VLAN for eth1(XORP) also? Am i doing anything wrong in configuring VLANS. in xorpsh > show pim interfces eth0 UP Sparse 2 DR 100 192.168.1.157 0 eth0.12 DISABLED Sparse 2 DR 1 10.0.3.3 0 eth0.45 DISABLED Sparse 2 DR 1 10.0.1.15 0 eth1 UP Sparse 2 DR 100 192.168.5.1 0 register_vif UP Sparse 2 DR 1 192.168.1.157 0 Thanks. Ramu [ 2006/06/05 19:45:24 INFO xorp_rtrmgr:24145 RTRMGR +99 module_manager.cc execute ] Executing module: fea (fea/xorp_fea) [ 2006/06/05 19:45:30 INFO xorp_rtrmgr:24145 RTRMGR +99 module_manager.cc execute ] Executing module: mfea4 (fea/xorp_fea) [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth0] pif_index: 2 vif_index: 0 addr: 192.168.1.157 subnet: 192.168.1.0/24broadcast: 192.168.1.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth0.12] pif_index: 5 vif_index: 1 addr: 10.0.3.3 subnet: 10.0.3.0/24 broadcast: 10.0.3.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth0.45] pif_index: 6 vif_index: 2 addr: 10.0.1.15 subnet: 10.0.1.0/24 broadcast: 10.0.1.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth1] pif_index: 3 vif_index: 3 addr: 192.168.5.1 subnet: 192.168.5.0/24broadcast: 192.168.5.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] MFEA started[ 2006/06/05 19:45:37 INFO xorp_rtrmgr:24145 RTRMGR +99 module_manager.cc execute ] Executing module: pimsm4 (pim/xorp_pimsm4) [ 2006/06/05 19:45:37 WARNING xorp_rtrmgr:24145 XrlFinderTarget +406 ../xrl/targets/finder_base.cc handle_finder_0_2_resolve_xrl ] Handling method for finder/0.2/resolve_xrl failed: XrlCmdError 102 Command failed Target "PIMSM_4" does not exist or is not enabled. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060606/6f4c8289/attachment.html From pavlin at icir.org Tue Jun 6 09:48:20 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Tue, 06 Jun 2006 09:48:20 -0700 Subject: [Xorp-users] VLAN Support in XORP In-Reply-To: Message from "Ramu k" of "Tue, 06 Jun 2006 21:32:32 +0530." Message-ID: <200606061648.k56GmKSn037727@possum.icir.org> > (I am working on IPTV testing. I am experimenting XORP for calculating join > latency > and leave latency. If i run one multicast user join latency is around (0.3msto > 0.6ms) > which quite OK . but leave latency i around 3secs which is very large > value.is it beacause > of XORP or its a problem with my machine(FC4).) How do you measure the leave latency? Is it the time between the moment you quit the receiver and the moment the multicast data stops flowing? In general, it takes on the order of 2-3 seconds (by default) for IGMP itself to declare that there is no other receiver on the subnet for the same multicast group. Hence, I believe that most of the 3 secs leave latency is because of the way IGMP operates. If you are really interested in decreasing the IGMP leave delay, there are some mechanisms for doing that, but currently this will be a bit tricky tricky (it may require editing of some of the XORP rtrmgr template files, etc), and there are some tradeoffs. Pavlin > I am facing a problem in the following scenario(VLANS). > My Lab setup is like below > > 192.168.5.7 (vlan eth0.12) > A----------------------XORP-----------------------------B > eth1 eth0 eth0 > 192.168.5.1 192.168.1.157 > > eth0.12 ip address is 10.0.3.3 > eth0.45 ip address is 10.0.3.15 > In Router machine(XORP) i configured 2 VLANS using command(eth0.12 ð0.45) > VCONFIG add eth0 12 . i added this VLAN to boot.config. > I am running mcast server on machine A(it has a VLAN 12 and ip 10.0.3.4). > i am getting packets to eth1 of XORP. but it is not forwarding eth0.12 > (ttl > 1 and added route in iptables also) > Do i need to have VLAN for eth1(XORP) also? > Am i doing anything wrong in configuring VLANS. > > > in xorpsh > show pim interfces > eth0 UP Sparse 2 DR 100 192.168.1.157 > 0 > eth0.12 DISABLED Sparse 2 DR 1 10.0.3.3 0 > eth0.45 DISABLED Sparse 2 DR 1 10.0.1.15 0 > eth1 UP Sparse 2 DR 100 192.168.5.1 > 0 > register_vif UP Sparse 2 DR 1 192.168.1.157 > 0 > > > Thanks. > Ramu > > > [ 2006/06/05 19:45:24 INFO xorp_rtrmgr:24145 RTRMGR +99 module_manager.cc > execute ] Executing module: fea (fea/xorp_fea) > [ 2006/06/05 19:45:30 INFO xorp_rtrmgr:24145 RTRMGR +99 module_manager.cc > execute ] Executing module: mfea4 (fea/xorp_fea) > [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth0] > pif_index: 2 vif_index: 0 addr: 192.168.1.157 subnet: 192.168.1.0/24broadcast: > 192.168.1.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP > [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth0.12] > pif_index: 5 vif_index: 1 addr: 10.0.3.3 subnet: 10.0.3.0/24 broadcast: > 10.0.3.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP > [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth0.45] > pif_index: 6 vif_index: 2 addr: 10.0.1.15 subnet: 10.0.1.0/24 broadcast: > 10.0.1.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP > [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] Interface added: Vif[eth1] > pif_index: 3 vif_index: 3 addr: 192.168.5.1 subnet: 192.168.5.0/24broadcast: > 192.168.5.255 peer: 0.0.0.0 Flags: MULTICAST BROADCAST UNDERLYING_VIF_UP > [ 2006/06/05 19:45:30 INFO xorp_fea MFEA ] MFEA started[ 2006/06/05 > 19:45:37 INFO xorp_rtrmgr:24145 RTRMGR +99 module_manager.cc execute ] > Executing module: pimsm4 (pim/xorp_pimsm4) > [ 2006/06/05 19:45:37 WARNING xorp_rtrmgr:24145 XrlFinderTarget +406 > ../xrl/targets/finder_base.cc handle_finder_0_2_resolve_xrl ] Handling > method for finder/0.2/resolve_xrl failed: XrlCmdError 102 Command failed > Target "PIMSM_4" does not exist or is not enabled. > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From becoxta at gmail.com Tue Jun 6 11:13:23 2006 From: becoxta at gmail.com (Bruno Alexandre D. da Costa / Grad. DSC) Date: Tue, 6 Jun 2006 15:13:23 -0300 Subject: [Xorp-users] Bidirectional PIM-SM Message-ID: Hi all I'm new in Xorp. I'm Testing two xorp routers talking PIM-SM each other with one of them as static RP, but it's not working. When visiting Roadmap xorp's page (http://www.xorp.org/roadmap.html) I see a Milestone Bidirectional PIM-SM. Have I to wait until september? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060606/de4e21fc/attachment.html From pavlin at icir.org Tue Jun 6 11:34:38 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Tue, 06 Jun 2006 11:34:38 -0700 Subject: [Xorp-users] Bidirectional PIM-SM In-Reply-To: Message from "Bruno Alexandre D. da Costa / Grad. DSC" of "Tue, 06 Jun 2006 15:13:23 -0300." Message-ID: <200606061834.k56IYcQd038804@possum.icir.org> > I'm new in Xorp. I'm Testing two xorp routers talking PIM-SM each other with > one of them as static RP, but it's not working. Please provide more details about the problem. E.g., config files, what exactly is working and not working, etc. > When visiting Roadmap xorp's page (http://www.xorp.org/roadmap.html) I see a > Milestone Bidirectional PIM-SM. > > Have I to wait until september? Yes, if you really need Bidir PIM-SM. Though, have in mind that priorities change over time so there is no guarantee it will be definitely implemented by then. Pavlin From lorenooliveira at gmail.com Wed Jun 7 04:37:59 2006 From: lorenooliveira at gmail.com (Loreno Oliveira) Date: Wed, 7 Jun 2006 08:37:59 -0300 Subject: [Xorp-users] XORP + VLC problem Message-ID: Hi folks! I?m on troubles with XORP + VLC. Me scenario is: I have a desktop in a wired network which is streaming video using VLC. A notebook running XORP has a wired and a wireless network cards, where the wired one is connected to the same network as the desktop and the wireless one is connected to another network. In the wireless side, I have a PDA, also running VLC, which tries to consume the multicast data streamed by the desktop. However, XORP is not forwarding the multicast traffic to the wireless interface. Does anybody has some hint about what is going wrong here?? Loreno -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060607/327db1a0/attachment.html From M.Handley at cs.ucl.ac.uk Wed Jun 7 06:33:43 2006 From: M.Handley at cs.ucl.ac.uk (Mark Handley) Date: Wed, 7 Jun 2006 14:33:43 +0100 Subject: [Xorp-users] XORP + VLC problem In-Reply-To: References: Message-ID: <84a612dd0606070633p260188edhf8471a2edf1acb25@mail.gmail.com> You'll need to give some more details. I assume you're configured to run PIM on the XORP notebook? Also that the PDA is successfully sending IGMP join messages that the XORP notebook can receive (otherwise it won't know there are any listeners). Can you send your xorp config file? - Mark On 6/7/06, Loreno Oliveira wrote: > Hi folks! > > I?m on troubles with XORP + VLC. Me scenario is: I have a desktop in a wired > network which is streaming video using VLC. A notebook running XORP has a > wired and a wireless network cards, where the wired one is connected to the > same network as the desktop and the wireless one is connected to another > network. > > In the wireless side, I have a PDA, also running VLC, which tries to consume > the multicast data streamed by the desktop. However, XORP is not forwarding > the multicast traffic to the wireless interface. > > > Does anybody has some hint about what is going wrong here?? > > Loreno > > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users > > > From vkaul at research.telcordia.com Wed Jun 7 08:17:39 2006 From: vkaul at research.telcordia.com (Vikram KAUL) Date: Wed, 7 Jun 2006 11:17:39 -0400 (Eastern Standard Time) Subject: [Xorp-users] ARP problems with static routes in XORP Message-ID: Folks, I need help with a skeletal XORP setup with 3 routers and two hosts. I am seeing the problem with static routes as well as with RIP, so I will just describe the static case, and hope I can resolve the problem The network is laid out like follows 17.17.17.x 10.1.1.x 192.168.0.x 16.16.16.x H ---------- R ---------- R ------------- R ----------- H .20 .1 .1 .2 .2 .3 .3 .20 HostA : 17.17.17.20 Router1: 17.17.17.1 & 10.1.1.1 Router2: 10.1.1.2 & 192.168.0.2 Router3: 192.168.0.3 & 16.16.16.3 HostB : 16.16.16.20 When I configure XORP for static routes (or even RIP), the unicast tables appear properly in the routers (1,2,3). I am able to ping and traceroute from one to the other without any problems Then, when I connect the host (HostA) and configure the default gw to be 17.17.17.1 it does the following: Ping from 17.17.17.20 (host) to 17.17.17.1 (gw router) is OK Ping from 17.17.17.20 (host) to 10.1.1.1 (gw router, next interface) is OK But Ping from 17.17.17.20 (host) to 10.1.1.2 (next hop after gw) FAILS Same is the case in the opposite direction I ran tcpdump on the 17.17.17.1 interface on Router1 and it is getting arp queries arp who-is 10.1.1.2 tell 17.17.17.20 Why is this happening ? Shouldn't the host merely send the ping packet to the gw router which should then forward it to the next hop and get the reply back in the same way ? Why is arp for 10.1.1.2 being sent out by the host (17.17.17.20) ? Is this a configuration issue at the host ? or something in XORP ? Thanks Vikram From allan at vyatta.com Wed Jun 7 08:35:03 2006 From: allan at vyatta.com (Allan Leinwand) Date: Wed, 7 Jun 2006 08:35:03 -0700 (PDT) Subject: [Xorp-users] ARP problems with static routes in XORP Message-ID: <3537519.151149694503950.JavaMail.root@mail.vyatta.com> Hi Vikram, It looks like your host may be doing the right thing, depending on how you have your default gateway set up. The host is sending an ARP for the destination IP address and the router (17.17.17.1) should reply with a proxy ARP if it has a route to 10.1.1.2. I'm assuming that this router has a directly connected route to 10.1.1.0/24. I would make sure that: 1) the host has a default gateway set to 17.17.17.1. You can use netstat -r (Linux/Win) and check for the presence of a 0.0.0.0/0 route to 17.17.17.1. 2) the router has a connected route to 10.1.1.0/24 (my assumption on your subnet mask) 3) the router at 10.1.1.2 has a route (static or RIP derived) back to 17.17.17.0/24 (again, I'm assuming you're using a /24 here again). Take care, Allan Vyatta, Inc. ----- Original Message ----- From: Vikram KAUL To: xorp-users at xorp.org Sent: Wednesday, June 7, 2006 8:17:39 AM GMT-0800 Subject: [Xorp-users] ARP problems with static routes in XORP Folks, I need help with a skeletal XORP setup with 3 routers and two hosts. I am seeing the problem with static routes as well as with RIP, so I will just describe the static case, and hope I can resolve the problem The network is laid out like follows 17.17.17.x 10.1.1.x 192.168.0.x 16.16.16.x H ---------- R ---------- R ------------- R ----------- H .20 .1 .1 .2 .2 .3 .3 .20 HostA : 17.17.17.20 Router1: 17.17.17.1 & 10.1.1.1 Router2: 10.1.1.2 & 192.168.0.2 Router3: 192.168.0.3 & 16.16.16.3 HostB : 16.16.16.20 When I configure XORP for static routes (or even RIP), the unicast tables appear properly in the routers (1,2,3). I am able to ping and traceroute from one to the other without any problems Then, when I connect the host (HostA) and configure the default gw to be 17.17.17.1 it does the following: Ping from 17.17.17.20 (host) to 17.17.17.1 (gw router) is OK Ping from 17.17.17.20 (host) to 10.1.1.1 (gw router, next interface) is OK But Ping from 17.17.17.20 (host) to 10.1.1.2 (next hop after gw) FAILS Same is the case in the opposite direction I ran tcpdump on the 17.17.17.1 interface on Router1 and it is getting arp queries arp who-is 10.1.1.2 tell 17.17.17.20 Why is this happening ? Shouldn't the host merely send the ping packet to the gw router which should then forward it to the next hop and get the reply back in the same way ? Why is arp for 10.1.1.2 being sent out by the host (17.17.17.20) ? Is this a configuration issue at the host ? or something in XORP ? Thanks Vikram _______________________________________________ Xorp-users mailing list Xorp-users at xorp.org http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From vkaul at research.telcordia.com Wed Jun 7 08:47:07 2006 From: vkaul at research.telcordia.com (Vikram KAUL) Date: Wed, 7 Jun 2006 11:47:07 -0400 (Eastern Standard Time) Subject: [Xorp-users] ARP problems with static routes in XORP In-Reply-To: <3537519.151149694503950.JavaMail.root@mail.vyatta.com> References: <3537519.151149694503950.JavaMail.root@mail.vyatta.com> Message-ID: Allan, Thanks for your reply allan> It looks like your host may be doing the right thing, depending on how you have your default gateway set up. The host is sending an ARP for the destination IP address and the router (17.17.17.1) should reply with a proxy ARP if it has a route to 10.1.1.2. I'm assuming that this router has a directly connected route to 10.1.1.0/24. The Router does get the arp request from the host, but there is no reply sent for that . I check that using tcpdump. Does something have to be enabled in XORP to reply arp requests for directly connected subnets ? I could not find that in the documentation. allan> allan>I would make sure that: allan>1) the host has a default gateway set to 17.17.17.1. You can use netstat -r (Linux/Win) and check for the presence of a 0.0.0.0/0 route to 17.17.17.1. allan>2) the router has a connected route to 10.1.1.0/24 (my assumption on your subnet mask) allan>3) the router at 10.1.1.2 has a route (static or RIP derived) back to 17.17.17.0/24 (again, I'm assuming you're using a /24 here again). I made sure of these when I was doing the test. Is there some hook/flag I can enable in XORP to see if the arps are being replied ? Thanks Vikram From lorenooliveira at gmail.com Wed Jun 7 08:49:37 2006 From: lorenooliveira at gmail.com (Loreno Oliveira) Date: Wed, 7 Jun 2006 12:49:37 -0300 Subject: [Xorp-users] XORP + VLC problem In-Reply-To: <84a612dd0606070633p260188edhf8471a2edf1acb25@mail.gmail.com> References: <84a612dd0606070633p260188edhf8471a2edf1acb25@mail.gmail.com> Message-ID: First of all, thanks to Mark and Carsten for the fast answers. Well, I'm afraid I've did a bad evaluation of where is the problem. In a test I?ve just finished, I monitored both network interfaces using tcpdump. What is really happening is: * XORP is started on notebook using config.boot below * on the desktop, I open VLC and start streaming a video * on the PDA, I open VLC and configure the multicast address for join * the video should start being reproduced now, but it doesn?t Well, taking a look at tcpdump output, what is happening is that the PDA sends the join message (tos 0xc0, ttl 1, id 0, offset 0, flags [DF], length: 32, optlength: 4 ( RA ) 192.168.10.25 > 224.1.1.224: igmp report 224.1.1.224). After that, several packages start being received in the wireless interface, where the PDA is in (tos 0x0, ttl 1, id 56963, offset 0, flags [none], length: 1356 xxx.xxx.63.228.1681 > 224.1.1.224.1234: UDP, length: 1328). The problem is that during the package reception, oddly, the PDA sends a leave message (tos 0xc0, ttl 1, id 0, offset 0, flags [DF], length: 32, optlength: 4 ( RA ) 192.168.10.25 > ALL-ROUTERS.MCAST.NET: igmp leave 224.1.1.224), therefore stoping the package forwarding. In such a scenario, it?s obvious that XORP is working properly, and VLC is somehow leaving the multicast group. Maybe VLC is analyzing the first packages and may be leaving the multicast group due to not supported codecs and/or data formats. Anyway, I think this list is not the place where to talk about VLC issues. At least I hope this thread be useful for people with similar problems. One more time, thanks for the answers. Loreno ========config.boot============ /* $XORP: xorp/rtrmgr/config.boot.sample,v 1.35 2006/03/09 12:18:20 pavlin Exp $ */ interfaces { restore-original-config-on-shutdown: false interface eth0 { description: "ethernet interface" disable: false default-system-config /*vif eth0 { disable: false address xxx.xxx.63.231 { prefix-length: 24 broadcast: xxx.xxx.63.255 disable: false } }*/ } interface pan0 { description: "bluetooth bridge" disable: false default-system-config /*vif pan0 { disable: false address 192.168.10.1 { prefix-length: 24 destination: 192.168.10.255 disable: false } }*/ } } fea { unicast-forwarding4 { disable: false } } plumbing { mfea4 { disable: false interface eth0 { vif eth0 { disable: false } } interface pan0 { vif pan0 { disable: false } } interface register_vif { vif register_vif { /* Note: this vif should be always enabled */ disable: false } } traceoptions { flag all { disable: false } } } } protocols { igmp { disable: false interface eth0 { vif eth0 { disable: false } } interface pan0 { vif pan0 { disable: false } } traceoptions { flag all { disable: false } } } } protocols { pimsm4 { disable: false interface eth0 { vif eth0 { disable: false dr-priority: 1 } } interface pan0 { vif pan0 { disable: false dr-priority: 1 } } interface register_vif { vif register_vif { /* Note: this vif should be always enabled */ disable: false } } static-rps { rp xxx.xxx.63.231 { group-prefix 224.0.0.0/4 { /* rp-priority: 192 */ /* hash-mask-len: 30 */ } } } /*bootstrap { disable: false cand-bsr { scope-zone 224.0.0.0/4 { cand-bsr-by-vif-name: "eth0" } } cand-rp { group-prefix 224.0.0.0/4 { cand-rp-by-vif-name: "eth0" } } }*/ switch-to-spt-threshold { /* approx. 1K bytes/s (10Kbps) threshold */ disable: false interval-sec: 100 bytes: 102400 } traceoptions { flag all { disable: false } } } } /* * Note: fib2mrib is needed for multicast only if the unicast protocols * don't populate the MRIB with multicast-specific routes. */ protocols { fib2mrib { disable: false } } On 6/7/06, Mark Handley wrote: > > You'll need to give some more details. > > I assume you're configured to run PIM on the XORP notebook? Also that > the PDA is successfully sending IGMP join messages that the XORP > notebook can receive (otherwise it won't know there are any > listeners). > > Can you send your xorp config file? > > - Mark > > On 6/7/06, Loreno Oliveira wrote: > > Hi folks! > > > > I?m on troubles with XORP + VLC. Me scenario is: I have a desktop in a > wired > > network which is streaming video using VLC. A notebook running XORP has > a > > wired and a wireless network cards, where the wired one is connected to > the > > same network as the desktop and the wireless one is connected to another > > network. > > > > In the wireless side, I have a PDA, also running VLC, which tries to > consume > > the multicast data streamed by the desktop. However, XORP is not > forwarding > > the multicast traffic to the wireless interface. > > > > > > Does anybody has some hint about what is going wrong here?? > > > > Loreno > > > > _______________________________________________ > > Xorp-users mailing list > > Xorp-users at xorp.org > > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060607/d066b547/attachment-0001.html From allan at vyatta.com Wed Jun 7 10:32:40 2006 From: allan at vyatta.com (Allan Leinwand) Date: Wed, 7 Jun 2006 10:32:40 -0700 (PDT) Subject: [Xorp-users] ARP problems with static routes in XORP Message-ID: <18785098.361149701560580.JavaMail.root@mail.vyatta.com> Hi Vikram, I believe that in XORP proxy ARP should be enabled by default. Are you sure that your routing table has the directly connected route to the subnet for 10.1.1.2? We've changed the routing table output in Vyatta OFR and I don't recall how to look at the unicast table in XORP. Our command is "show route" :) Take care, Allan ----- Original Message ----- From: Vikram KAUL To: Allan Leinwand Cc: xorp-users at xorp.org Sent: Wednesday, June 7, 2006 8:47:07 AM GMT-0800 Subject: Re: [Xorp-users] ARP problems with static routes in XORP Allan, Thanks for your reply allan> It looks like your host may be doing the right thing, depending on how you have your default gateway set up. The host is sending an ARP for the destination IP address and the router (17.17.17.1) should reply with a proxy ARP if it has a route to 10.1.1.2. I'm assuming that this router has a directly connected route to 10.1.1.0/24. The Router does get the arp request from the host, but there is no reply sent for that . I check that using tcpdump. Does something have to be enabled in XORP to reply arp requests for directly connected subnets ? I could not find that in the documentation. allan> allan>I would make sure that: allan>1) the host has a default gateway set to 17.17.17.1. You can use netstat -r (Linux/Win) and check for the presence of a 0.0.0.0/0 route to 17.17.17.1. allan>2) the router has a connected route to 10.1.1.0/24 (my assumption on your subnet mask) allan>3) the router at 10.1.1.2 has a route (static or RIP derived) back to 17.17.17.0/24 (again, I'm assuming you're using a /24 here again). I made sure of these when I was doing the test. Is there some hook/flag I can enable in XORP to see if the arps are being replied ? Thanks Vikram _______________________________________________ Xorp-users mailing list Xorp-users at xorp.org http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From mhorn at vyatta.com Wed Jun 7 10:37:46 2006 From: mhorn at vyatta.com (Mike Horn) Date: Wed, 7 Jun 2006 11:37:46 -0600 Subject: [Xorp-users] ARP problems with static routes in XORP In-Reply-To: Message-ID: <11f901c68a59$189fdc10$0f02a8c0@caddisconsulting.com> Hi Vikram, To help you troubleshoot, can you provide a few pieces of information: 1. What OS are the hosts and Routers? 2. Can you send the interface configuration from all the devices (ifconfig or equivalent). 3. Can you send the routing table output from all the devices. You can use "netstat -rn" from the unix shell or you can use XORP command "show route table ipv4 unicast final". Proxy ARP won't work in your setup, the HostA should only be ARP'ing for the nexthop address on the local connected subnet, in this case 17.17.17.1. The traffic bound for 10.1.1.2 should be sent to 17.17.17.1 who should then forward it to 10.1.1.2. -mike -----Original Message----- From: xorp-users-bounces at xorp.org [mailto:xorp-users-bounces at xorp.org] On Behalf Of Vikram KAUL Sent: Wednesday, June 07, 2006 9:47 AM To: Allan Leinwand Cc: xorp-users at xorp.org Subject: Re: [Xorp-users] ARP problems with static routes in XORP Allan, Thanks for your reply allan> It looks like your host may be doing the right thing, depending on how you have your default gateway set up. The host is sending an ARP for the destination IP address and the router (17.17.17.1) should reply with a proxy ARP if it has a route to 10.1.1.2. I'm assuming that this router has a directly connected route to 10.1.1.0/24. The Router does get the arp request from the host, but there is no reply sent for that . I check that using tcpdump. Does something have to be enabled in XORP to reply arp requests for directly connected subnets ? I could not find that in the documentation. allan> allan>I would make sure that: allan>1) the host has a default gateway set to 17.17.17.1. You can use netstat -r (Linux/Win) and check for the presence of a 0.0.0.0/0 route to 17.17.17.1. allan>2) the router has a connected route to 10.1.1.0/24 (my assumption allan>on your subnet mask) allan>3) the router at 10.1.1.2 has a route (static or RIP derived) back to 17.17.17.0/24 (again, I'm assuming you're using a /24 here again). I made sure of these when I was doing the test. Is there some hook/flag I can enable in XORP to see if the arps are being replied ? Thanks Vikram _______________________________________________ Xorp-users mailing list Xorp-users at xorp.org http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From vkaul at research.telcordia.com Wed Jun 7 10:47:05 2006 From: vkaul at research.telcordia.com (Vikram KAUL) Date: Wed, 7 Jun 2006 13:47:05 -0400 (Eastern Standard Time) Subject: [Xorp-users] ARP problems with static routes in XORP In-Reply-To: <11f901c68a59$189fdc10$0f02a8c0@caddisconsulting.com> References: <11f901c68a59$189fdc10$0f02a8c0@caddisconsulting.com> Message-ID: mhorn>To help you troubleshoot, can you provide a few pieces of information: mhorn>1. What OS are the hosts and Routers? mhorn>2. Can you send the interface configuration from all the devices (ifconfig mhorn>or equivalent). mhorn>3. Can you send the routing table output from all the devices. You can use mhorn>"netstat -rn" from the unix shell or you can use XORP command "show route mhorn>table ipv4 unicast final". I will be able to do this tomorrow. I am in another location now. mhorn> mhorn>Proxy ARP won't work in your setup, the HostA should only be ARP'ing for the mhorn>nexthop address on the local connected subnet, in this case 17.17.17.1. The mhorn>traffic bound for 10.1.1.2 should be sent to 17.17.17.1 who should then mhorn>forward it to 10.1.1.2. mhorn> That is what I thought should happen. The fact that 10.1.1.2 is corresponding to the default entry of the route table at the host for which the gatway is 17.17.17.1 should make the host send a ping directly to the gatway. The gateway router would then forward the ping to the next hop router. I will send the XORP configurations tomorrow. In the meanwhile, if anyone has any other recommendations, please do advise. regards.. Vikram From allan at vyatta.com Wed Jun 7 11:31:01 2006 From: allan at vyatta.com (Allan Leinwand) Date: Wed, 7 Jun 2006 11:31:01 -0700 (PDT) Subject: [Xorp-users] ARP problems with static routes in XORP Message-ID: <22077272.451149705061293.JavaMail.root@mail.vyatta.com> Hi Vikram, It's best to go with advice from Mike Horn over mine everyday. He's a great technical resource that has been contributing to our project... I'll go back to ssuming my role in management now and try to forget that I was once technical :) :) Take care, Allan Vyatta, Inc. ----- Original Message ----- From: Vikram KAUL To: Mike Horn Cc: Allan Leinwand , xorp-users at xorp.org Sent: Wednesday, June 7, 2006 10:47:05 AM GMT-0800 Subject: RE: [Xorp-users] ARP problems with static routes in XORP mhorn>To help you troubleshoot, can you provide a few pieces of information: mhorn>1. What OS are the hosts and Routers? mhorn>2. Can you send the interface configuration from all the devices (ifconfig mhorn>or equivalent). mhorn>3. Can you send the routing table output from all the devices. You can use mhorn>"netstat -rn" from the unix shell or you can use XORP command "show route mhorn>table ipv4 unicast final". I will be able to do this tomorrow. I am in another location now. mhorn> mhorn>Proxy ARP won't work in your setup, the HostA should only be ARP'ing for the mhorn>nexthop address on the local connected subnet, in this case 17.17.17.1. The mhorn>traffic bound for 10.1.1.2 should be sent to 17.17.17.1 who should then mhorn>forward it to 10.1.1.2. mhorn> That is what I thought should happen. The fact that 10.1.1.2 is corresponding to the default entry of the route table at the host for which the gatway is 17.17.17.1 should make the host send a ping directly to the gatway. The gateway router would then forward the ping to the next hop router. I will send the XORP configurations tomorrow. In the meanwhile, if anyone has any other recommendations, please do advise. regards.. Vikram From M.Handley at cs.ucl.ac.uk Wed Jun 7 13:04:58 2006 From: M.Handley at cs.ucl.ac.uk (Mark Handley) Date: Wed, 7 Jun 2006 21:04:58 +0100 Subject: [Xorp-users] ARP problems with static routes in XORP In-Reply-To: References: Message-ID: <84a612dd0606071304w715c7f3as48cf3e716bb75fd5@mail.gmail.com> On 6/7/06, Vikram KAUL wrote: > I ran tcpdump on the 17.17.17.1 interface on Router1 and it is getting arp > queries > arp who-is 10.1.1.2 tell 17.17.17.20 > > Why is this happening ? The host should not be arping for an address that is not on the local subnet. My guess is that the host 17.17.17.20 thinks that 10.1.1.2 is in fact on its local subnet. The obvious cause for this would be that the netmask on the host is set incorrectly. Of course it could be something else, but this is the simplest explanation that fits the symptoms. - Mark From daniel at jerveren.se Thu Jun 8 02:52:51 2006 From: daniel at jerveren.se (Daniel) Date: Thu, 08 Jun 2006 11:52:51 +0200 Subject: [Xorp-users] Xorp and SNMP on Fedora Core 5 Message-ID: <4487F373.8080501@jerveren.se> I'm getting errors when trying to configure xorp with SNMP support on a Fedora system, ./configure --with-snmp gives me: ... checking whether C compiler supports flag "-DINET6 -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m32 -march=i386 -mtune=generic -fasynchronous-unwind-tables -lcrypto -Dlinux -I/usr/include/rpm -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -pipe -Wdeclaration-after-statement -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm -I/usr/lib/perl5/5.8.8/i386-linux-thread-multi/CORE -I. -I/usr/include/net-snmp -DNETSNMP_NO_INLINE"... ./configure: line 5226: ${ac_cv_prog_c_compiler__DINET6__O2__g__pipe__Wall__Wp,_D_FORTIFY_SOURCE=2__fexceptions__fstack_protector___param=ssp_buffer_size=4__m32__march=i386__mtune=generic__fasynchronous_unwind_tables__lcrypto__Dlinux__I_usr_include_rpm__D_REENTRANT__D_GNU_SOURCE__fno_strict_aliasing__pipe__Wdeclaration_after_statement__I_usr_local_include__D_LARGEFILE_SOURCE__D_FILE_OFFSET_BITS=64__I_usr_include_gdbm__I_usr_lib_perl5_5_8_8_i386_linux_thread_multi_CORE__I___I_usr_include_net_snmp__DNETSNMP_NO_INLINE+set}: bad substitution ./configure: line 5252: ac_cv_prog_c_compiler__DINET6__O2__g__pipe__Wall__Wp,_D_FORTIFY_SOURCE=2__fexceptions__fstack_protector___param=ssp_buffer_size=4__m32__march=i386__mtune=generic__fasynchronous_unwind_tables__lcrypto__Dlinux__I_usr_include_rpm__D_REENTRANT__D_GNU_SOURCE__fno_strict_aliasing__pipe__Wdeclaration_after_statement__I_usr_local_include__D_LARGEFILE_SOURCE__D_FILE_OFFSET_BITS=64__I_usr_include_gdbm__I_usr_lib_perl5_5_8_8_i386_linux_thread_multi_CORE__I___I_usr_include_net_snmp__DNETSNMP_NO_INLINE=yes: command not found no ------------------------------------------------------------- 'net-snmp-config --cflags' gave invalid cflags! ------------------------------------------------------------- configure: WARNING: XORP MIBs will not be built ... I've tried Xorp 1.2 and todays CVS, both give me the same error. The following relevant(?) software is installed on the system: gcc version 4.1.1 libtool 1.5.22 GNU Make 3.80 autoconf 2.59 automake 1.9.6 net-snmp-5.3-4.2 net-snmp-utils-5.3-4.2 net-snmp-libs-5.3-4.2 net-snmp-devel-5.3-4.2 Any suggestions? Best regards /d'. From orante2003 at yahoo.it Thu Jun 8 03:49:08 2006 From: orante2003 at yahoo.it (ORANTE TUCCERI) Date: Thu, 8 Jun 2006 12:49:08 +0200 (CEST) Subject: [Xorp-users] xorp_measures Message-ID: <20060608104908.92442.qmail@web26805.mail.ukl.yahoo.com> hello to all, I am executing of the measures on the software router xorp. The measures have the scope to estimate the performances of the router in terms of time demanded for the execution of the calculation of the shortest path tree. From the measures conduct on topologies of net (emulated through PC) fully connected constituted from a number of router R 10: 10: 100 (min /step /max) with interposed of the transit network between router and an other, is possible to obtain the following course (Time to calculate the SPT in milli-seconds vs.Vertex N) N= R*(R-1)*0.5 Xorp -1.2 vs. Quagga 0.98 not optimized see the attached graph We think that the evidenced not linear behavior from the diagram is legacy to the choice of the given structures associates, second given definition into RFC 2328, to the list of the candidates, the list of the next-hop and to the database of the LSA. I ask confirmation for following associations XORP/RFC2328 -----xorp/libproto/spt.hh @726 _ tentative/list of the candidates -----xorp/libproto/area_router.hh @586 vector _ db// Database of LSAs Other aspect that me is not clear is this: If the association _ db/Database of the LSARef is corrected, ? since LsaRef is a alias of ref_ptr ? since the class ref_ptr (present in libxorp/ref_ptr.hh) is created with the scope to maintain the count of the number of how many references exists to an object me it is not clearly because the memory of the database of the LSA is implemented through carrier of type objects '' ref_ptr ' '. tanks for attention. < orante2003 at yahoo.it > Chiacchiera con i tuoi amici in tempo reale! http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060608/6a7fadf6/attachment-0001.html -------------- next part -------------- A non-text attachment was scrubbed... Name: Provexuna .jpg Type: image/pjpeg Size: 24040 bytes Desc: 1115954615-Provexuna .jpg Url : http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060608/6a7fadf6/attachment-0001.bin From M.Handley at cs.ucl.ac.uk Thu Jun 8 05:17:17 2006 From: M.Handley at cs.ucl.ac.uk (Mark Handley) Date: Thu, 8 Jun 2006 13:17:17 +0100 Subject: [Xorp-users] xorp_measures In-Reply-To: <20060608104908.92442.qmail@web26805.mail.ukl.yahoo.com> References: <20060608104908.92442.qmail@web26805.mail.ukl.yahoo.com> Message-ID: <84a612dd0606080517v3da4a5bareae5eab587373e9d@mail.gmail.com> >From looking at the graph, it looks like the behaviour may in fact be linear in the limit. The three highest data points appear to lie on a line (although it's difficult to draw a strong conclusion from such limited data). Some more data points in this region would be useful. However, if it is linear for large N, then this doesn't make much sense to me, as extrapolating this line back intercepts x=0 at a negative value. What does it look like on a log-log scale? - Mark On 6/8/06, ORANTE TUCCERI wrote: > > > hello to all, > I am executing of the measures on the software router xorp. > The measures have the scope to estimate the performances of the router in > terms > of time demanded for the execution of the calculation of the shortest path > tree. > From the measures conduct on topologies of net (emulated through PC) fully > connected constituted from a number of router > R 10: 10: 100 (min /step /max) > with interposed of the transit network between router and an other, is > possible > to obtain the following course > (Time to calculate the SPT in milli-seconds vs.Vertex N) > > N= R*(R-1)*0.5 > Xorp -1.2 vs. Quagga 0.98 not optimized see the attached graph > > We > think that the evidenced not linear behavior from the diagram is legacy to > the > choice of the given structures associates, second given definition into RFC > 2328, to the list of the candidates, the list of the next-hop and to the > database of the LSA. > I ask confirmation for following associations XORP/RFC2328 > -----xorp/libproto/spt.hh > @726 _ tentative/list of the candidates > -----xorp/libpproto/area_router.hh > @586 vector _ db// Database of LSAs > Other aspect that me is not clear is this: > If the association > _ db/Database of the LSARef is corrected, > ? since LsaRef is a alias of ref_ptr > ? since the class ref_ptr (present in libxorp/ref_ptr.hh) is created with > the scope to maintain the count of the number of how many references exists > to an object > me it is not clearly because the memory of the database of the LSA is > implemented through carrier of type objects '' ref_ptr ' '. > tanks for attention. > < orante2003 at yahoo.it > > > Chiacchiera con i tuoi amici in tempo reale! > http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users > > > > From pavlin at icir.org Thu Jun 8 16:22:21 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Thu, 08 Jun 2006 16:22:21 -0700 Subject: [Xorp-users] Xorp and SNMP on Fedora Core 5 In-Reply-To: Message from Daniel of "Thu, 08 Jun 2006 11:52:51 +0200." <4487F373.8080501@jerveren.se> Message-ID: <200606082323.k58NMLkm046879@possum.icir.org> Daniel wrote: > I'm getting errors when trying to configure xorp with SNMP support on a > Fedora system, ./configure --with-snmp gives me: Daniel, What is the output of "net-snmp-config --cflags" and "net-snmp-config --version" Pavlin > > > ... > checking whether C compiler supports flag "-DINET6 -O2 -g -pipe -Wall > -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector > --param=ssp-buffer-size=4 -m32 -march=i386 -mtune=generic > -fasynchronous-unwind-tables -lcrypto -Dlinux -I/usr/include/rpm > -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -pipe > -Wdeclaration-after-statement -I/usr/local/include -D_LARGEFILE_SOURCE > -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm > -I/usr/lib/perl5/5.8.8/i386-linux-thread-multi/CORE -I. > -I/usr/include/net-snmp -DNETSNMP_NO_INLINE"... ./configure: line 5226: > ${ac_cv_prog_c_compiler__DINET6__O2__g__pipe__Wall__Wp,_D_FORTIFY_SOURCE=2__fexceptions__fstack_protector___param=ssp_buffer_size=4__m32__march=i386__mtune=generic__fasynchronous_unwind_tables__lcrypto__Dlinux__I_usr_include_rpm__D_REENTRANT__D_GNU_SOURCE__fno_strict_aliasing__pipe__Wdeclaration_after_statement__I_usr_local_include__D_LARGEFILE_SOURCE__D_FILE_OFFSET_BITS=64__I_usr_include_gdbm__I_usr_lib_perl5_5_8_8_i386_linux_thread_multi_CORE__I___I_usr_include_net_snmp__DNETSNMP_NO_INLINE+set}: > bad substitution > ./configure: line 5252: > ac_cv_prog_c_compiler__DINET6__O2__g__pipe__Wall__Wp,_D_FORTIFY_SOURCE=2__fexceptions__fstack_protector___param=ssp_buffer_size=4__m32__march=i386__mtune=generic__fasynchronous_unwind_tables__lcrypto__Dlinux__I_usr_include_rpm__D_REENTRANT__D_GNU_SOURCE__fno_strict_aliasing__pipe__Wdeclaration_after_statement__I_usr_local_include__D_LARGEFILE_SOURCE__D_FILE_OFFSET_BITS=64__I_usr_include_gdbm__I_usr_lib_perl5_5_8_8_i386_linux_thread_multi_CORE__I___I_usr_include_net_snmp__DNETSNMP_NO_INLINE=yes: > command not found > no > ------------------------------------------------------------- > 'net-snmp-config --cflags' gave invalid cflags! > ------------------------------------------------------------- > configure: WARNING: XORP MIBs will not be built > ... > > > > I've tried Xorp 1.2 and todays CVS, both give me the same error. The > following relevant(?) software is installed on the system: > > gcc version 4.1.1 > libtool 1.5.22 > GNU Make 3.80 > autoconf 2.59 > automake 1.9.6 > net-snmp-5.3-4.2 > net-snmp-utils-5.3-4.2 > net-snmp-libs-5.3-4.2 > net-snmp-devel-5.3-4.2 > > Any suggestions? > > Best regards /d'. > > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From daniel at jerveren.se Thu Jun 8 22:15:40 2006 From: daniel at jerveren.se (Daniel) Date: Fri, 09 Jun 2006 07:15:40 +0200 Subject: [Xorp-users] Xorp and SNMP on Fedora Core 5 In-Reply-To: <200606082323.k58NMLkm046879@possum.icir.org> References: <200606082323.k58NMLkm046879@possum.icir.org> Message-ID: <448903FC.9000803@jerveren.se> Pavlin Radoslavov wrote: > What is the output of "net-snmp-config --cflags" and > "net-snmp-config --version" [daniel at cnode2 xorp]$ net-snmp-config --cflags -DINET6 -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m32 -march=i386 -mtune=generic -fasynchronous-unwind-tables -lcrypto -Dlinux -I/usr/include/rpm -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -pipe -Wdeclaration-after-statement -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm -I/usr/lib/perl5/5.8.8/i386-linux-thread-multi/CORE -I. -I/usr/include/net-snmp [daniel at cnode2 xorp]$ net-snmp-config --version 5.3 All the best, /d'. From pavlin at icir.org Thu Jun 8 23:45:22 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Thu, 08 Jun 2006 23:45:22 -0700 Subject: [Xorp-users] Xorp and SNMP on Fedora Core 5 In-Reply-To: Message from Daniel of "Fri, 09 Jun 2006 07:15:40 +0200." <448903FC.9000803@jerveren.se> Message-ID: <200606090645.k596jM4S084212@possum.icir.org> > Pavlin Radoslavov wrote: > > > What is the output of "net-snmp-config --cflags" and > > "net-snmp-config --version" > > [daniel at cnode2 xorp]$ net-snmp-config --cflags > -DINET6 -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions > -fstack-protector --param=ssp-buffer-size=4 -m32 -march=i386 > -mtune=generic -fasynchronous-unwind-tables -lcrypto -Dlinux > -I/usr/include/rpm -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -pipe > -Wdeclaration-after-statement -I/usr/local/include -D_LARGEFILE_SOURCE > -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm > -I/usr/lib/perl5/5.8.8/i386-linux-thread-multi/CORE -I. > -I/usr/include/net-snmp > > [daniel at cnode2 xorp]$ net-snmp-config --version > 5.3 Thanks for the info. The problem is now fixed in CVS: Revision Changes Path 1.9 +15 -19; commitid: 17eb6448916687ea6; xorp/config/compiler_flags.m4 1.195 +9 -63; commitid: 17eb6448916687ea6; xorp/configure 1.34 +10 -70; commitid: 17eb6448916687ea6; xorp/mibs/configure Pavlin From atanu at ICSI.Berkeley.EDU Fri Jun 9 14:40:10 2006 From: atanu at ICSI.Berkeley.EDU (Atanu Ghosh) Date: Fri, 09 Jun 2006 14:40:10 -0700 Subject: [Xorp-users] xorp_measures In-Reply-To: Message from ORANTE TUCCERI of "Thu, 08 Jun 2006 12:49:08 +0200." <20060608104908.92442.qmail@web26805.mail.ukl.yahoo.com> Message-ID: <31391.1149889210@tigger.icir.org> Hi, I don't understand your experiment but if the graphs that are being tested are fully connected then for every additional vertex {vertex - 1} edges will be added (in the 5000 vertex case each vertex will have 4999 edges), will any algorithm that has to traverse all edges perform linearly as the vertexes are increased? The SPT code in XORP is in a library and independant of OSFP the interface to the library was designed to allow the implementation to use an incremental algorithm. After we measured the performance of the SPT code we didn't feel that it was necessary to add the incremental support yet, for you experiment am incremental implementation would probably have run slower as every vertex is modified. I don't understand your questions about the code. It seems unlikely that the choice of data structures in OSPF will make much difference to the SPT computation time which is in a separate library. For very LSA there is a single reference counted copy, to simplify management. Atanu. >>>>> "ORANTE" == ORANTE TUCCERI writes: ORANTE> hello to all, I am executing of the measures on the ORANTE> software router xorp. The measures have the scope to ORANTE> estimate the performances of the router in terms of time ORANTE> demanded for the execution of the calculation of the ORANTE> shortest path tree. From the measures conduct on topologies ORANTE> of net (emulated through PC) fully connected constituted ORANTE> from a number of router R 10: 10: 100 (min /step /max) with ORANTE> interposed of the transit network between router and an ORANTE> other, is possible to obtain the following course (Time to ORANTE> calculate the SPT in milli-seconds vs.Vertex N) ORANTE> N= R*(R-1)*0.5 Xorp -1.2 vs. Quagga 0.98 not optimized ORANTE> see the attached graph ORANTE> We think that the evidenced not linear behavior from the ORANTE> diagram is legacy to the choice of the given structures ORANTE> associates, second given definition into RFC 2328, to the ORANTE> list of the candidates, the list of the next-hop and to the ORANTE> database of the LSA. I ask confirmation for following ORANTE> associations XORP/RFC2328 -----xorp/libproto/spt.hh @726 _ ORANTE> tentative/list of the candidates ORANTE> -----xorp/libproto/area_router.hh @586 vector _ db// ORANTE> Database of LSAs Other aspect that me is not clear is this: ORANTE> If the association _ db/Database of the LSARef is corrected, ORANTE> ? since LsaRef is a alias of ref_ptr ? since the class ORANTE> ref_ptr (present in libxorp/ref_ptr.hh) is created with the ORANTE> scope to maintain the count of the number of how many ORANTE> references exists to an object me it is not clearly because ORANTE> the memory of the database of the LSA is implemented through ORANTE> carrier of type objects '' ref_ptr ' '. tanks for ORANTE> attention. < orante2003 at yahoo.it > ORANTE> Chiacchiera con i tuoi amici in tempo reale! ORANTE> http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com ORANTE> _______________________________________________ Xorp-users ORANTE> mailing list Xorp-users at xorp.org ORANTE> http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From vkaul at research.telcordia.com Mon Jun 12 10:15:41 2006 From: vkaul at research.telcordia.com (Vikram KAUL) Date: Mon, 12 Jun 2006 13:15:41 -0400 (Eastern Standard Time) Subject: [Xorp-users] ARP problems with static routes in XORP In-Reply-To: References: <11f901c68a59$189fdc10$0f02a8c0@caddisconsulting.com> Message-ID: Mike, Allan I confirmed that the error was due to some 'funky' routes on the host, and nothing was wrong in the XORP configuration of static routes. The ping works now. I also ran rip instead of using static routes, and the routes get created and ping works fine Thanks for your help Vikram vkaul>Date: Wed, 7 Jun 2006 13:47:05 -0400 (Eastern Standard Time) vkaul>To: Mike Horn vkaul>Cc: xorp-users at xorp.org, 'Allan Leinwand' vkaul>Subject: Re: [Xorp-users] ARP problems with static routes in XORP vkaul> vkaul> vkaul>mhorn>To help you troubleshoot, can you provide a few pieces of information: vkaul>mhorn>1. What OS are the hosts and Routers? vkaul>mhorn>2. Can you send the interface configuration from all the devices (ifconfig vkaul>mhorn>or equivalent). vkaul>mhorn>3. Can you send the routing table output from all the devices. You can use vkaul>mhorn>"netstat -rn" from the unix shell or you can use XORP command "show route vkaul>mhorn>table ipv4 unicast final". vkaul> vkaul>I will be able to do this tomorrow. I am in another location now. vkaul> vkaul>mhorn> vkaul>mhorn>Proxy ARP won't work in your setup, the HostA should only be ARP'ing for the vkaul>mhorn>nexthop address on the local connected subnet, in this case 17.17.17.1. The vkaul>mhorn>traffic bound for 10.1.1.2 should be sent to 17.17.17.1 who should then vkaul>mhorn>forward it to 10.1.1.2. vkaul>mhorn> vkaul> vkaul>That is what I thought should happen. The fact that 10.1.1.2 is vkaul>corresponding to the default entry of the route table at the host for vkaul>which the gatway is 17.17.17.1 should make the host send a ping directly vkaul>to the gatway. The gateway router would then forward the ping to the next vkaul>hop router. vkaul> vkaul>I will send the XORP configurations tomorrow. In the meanwhile, if anyone vkaul>has any other recommendations, please do advise. vkaul> vkaul>regards.. vkaul>Vikram vkaul> vkaul>_______________________________________________ vkaul>Xorp-users mailing list vkaul>Xorp-users at xorp.org vkaul>http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users vkaul> From venkthi1 at iit.edu Thu Jun 22 00:42:15 2006 From: venkthi1 at iit.edu (Venketesan) Date: Thu, 22 Jun 2006 00:42:15 -0700 Subject: [Xorp-users] IGMPv3 Status/ v1.3 release Message-ID: <2461ce524605e1.24605e12461ce5@iit.edu> Hi, Could someone from the dev team update about a tentative release of XORP 1.3 with IGMPv3 support. Once this is provided, does it mean that we have complete SSM support for IPv4. I mean can we run SSM traffic on a network of XORP 1.3 routers. I believe if the leaf multicast router supports IGMPv3 and implements PIM-SM2 -v11-ietf draft (currenrly avalialbe in XORP), then (S,G) Joins can be directly initated by the leaf router without first trying to build a RP tree. Hence the setup of an SSM multicat channel. Please confirm. regards, venkat From pavlin at icir.org Thu Jun 22 00:49:53 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Thu, 22 Jun 2006 00:49:53 -0700 Subject: [Xorp-users] IGMPv3 Status/ v1.3 release In-Reply-To: Message from Venketesan of "Thu, 22 Jun 2006 00:42:15 PDT." <2461ce524605e1.24605e12461ce5@iit.edu> Message-ID: <200606220749.k5M7nrfX007858@possum.icir.org> > Could someone from the dev team update about a tentative release of > XORP 1.3 with IGMPv3 support. The IGMPv3 implementation is work in progress as we speak. It should be ready for the next 1.3 release. > Once this is provided, does it mean that we have complete SSM support > for IPv4. I mean can we run SSM traffic on a network of XORP 1.3 > routers. Yes. > I believe if the leaf multicast router supports IGMPv3 and implements > PIM-SM2 -v11-ietf draft (currenrly avalialbe in XORP), then (S,G) > Joins can be directly initated by the leaf router without first trying > to build a RP tree. Hence the setup of an SSM multicat channel. Please > confirm. Yes. Pavlin From venkthi1 at iit.edu Thu Jun 22 07:42:02 2006 From: venkthi1 at iit.edu (Venketesan) Date: Thu, 22 Jun 2006 07:42:02 -0700 Subject: [Xorp-users] IGMPv3 Status/ v1.3 release Message-ID: <24745062471a96.2471a962474506@iit.edu> Thanks for the clarification. As i see in the road map that V1.3 is slated to release in June 2006. Could give an idea around what time/delay would it be released. regads, Venkat ----- Original Message ----- From: Pavlin Radoslavov Date: Thursday, June 22, 2006 0:49 am Subject: Re: [Xorp-users] IGMPv3 Status/ v1.3 release > > Could someone from the dev team update about a tentative release > of > > XORP 1.3 with IGMPv3 support. > > The IGMPv3 implementation is work in progress as we speak. > It should be ready for the next 1.3 release. > > > Once this is provided, does it mean that we have complete SSM > support > > for IPv4. I mean can we run SSM traffic on a network of XORP 1.3 > > routers. > > Yes. > > > I believe if the leaf multicast router supports IGMPv3 and > implements > > PIM-SM2 -v11-ietf draft (currenrly avalialbe in XORP), then > (S,G) > > Joins can be directly initated by the leaf router without first > trying > > to build a RP tree. Hence the setup of an SSM multicat channel. > Please > > confirm. > > Yes. > > Pavlin > From pavlin at icir.org Thu Jun 22 08:32:18 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Thu, 22 Jun 2006 08:32:18 -0700 Subject: [Xorp-users] IGMPv3 Status/ v1.3 release In-Reply-To: Message from Venketesan of "Thu, 22 Jun 2006 07:42:02 PDT." <24745062471a96.2471a962474506@iit.edu> Message-ID: <200606221532.k5MFWI09017875@possum.icir.org> > Thanks for the clarification. As i see in the road map that V1.3 is > slated to release in June 2006. Could give an idea around what > time/delay would it be released. Currently, we are trying to get the release candidate by the end of June (or beginning of July if things get delayed). The release itself will be 2 weeks after that. Pavlin From morekonda at gmail.com Thu Jun 22 22:46:11 2006 From: morekonda at gmail.com (Ganges) Date: Fri, 23 Jun 2006 11:16:11 +0530 Subject: [Xorp-users] morekonda@gmail.com Message-ID: Hi, I'm new to the XORP community. I have tried to download the LiveCD Rel. 1.2on Windows XP machine, it was downloaded with extenson .gz and surprisingly the file size was about 254MB instead of 89MB. I had tough time to burn the CD. Later i tried the same with Linux machine and somehow got the proper .iso image and able to burn on the CD. I tried to run the LiveCD image from a PC which, has latest chipsets, motheboard and interfaces. It worked until the XOPR splash image and hung. Appreciate any help regarding any hardware configuration/specification and download procedure would really help me to proceed further. Look forward for any valuable help. -- Cheers, Ganges.... -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060623/bd9b40ec/attachment.html From solca at guug.org Fri Jun 23 01:12:55 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 03:12:55 -0500 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup Message-ID: <20060623081255.GC22337@guug.org> Hi again again! :) I have this complex setup in Debian sarge for my University (In Guatemala so no budget for Cisco/Juniper routers): If IP address Purpose -- ---------- ------- eth0 10.0.0.1/24 Internal LAN eth1 10.0.1.1/24 Internal LAN eth2 172.16.0.1/16 Internal LAN eth3 192.168.0.1/24 Internal LAN eth3.2 192.168.2.1/24 Internal LAN (VLAN) eth3.3 192.168.3.1/24 Internal LAN (VLAN) eth3.4 192.168.4.1/24 Internal LAN (VLAN) eth3.5 192.168.5.1/24 Internal LAN (VLAN) eth3.6 192.168.6.1/24 Internal LAN (VLAN) eth3.7 192.168.7.1/24 Internal LAN (VLAN) eth3.8 192.168.8.1/24 Internal LAN (VLAN) eth3.9 192.168.9.1/24 Internal LAN (VLAN) eth3.10 192.168.10.1/24 Internal LAN (VLAN) eth3.15 192.168.15.1/24 Internal LAN (VLAN) eth4 10.0.2.1/24 Internal LAN eth5 (unused) eth6 10.10.26.7/28 External to Internet2 eth7 * External to Internet * eth7 have 120+ assigned IPv4 address for NATing internal hosts. I am using Xorp for 3 purposes: 1. BGP/mBGP with peer in eth6 for connecting to Internet2. 2. IGMP tracking for my internal hosts. 3. PIM-SM (multicast) for my internal hosts and for Internet2. BGP/mBGP works, I obtain more than 9000 routes. IGMP and PIM-SM in my internal LANs works as expected, I can stream to a multicast address and any host from any internal LAN can 'see' the stream if he join it. Now my problems: a. My next-hop router for Internet2 is Linux too and there is no MSDP implementation for Linux I choose to use the next-next router (Cisco) as my RP in PIM-SM. But when sniffing the wire my box is sending PIM-register to the Cisco (via eth6) but with the source-address as 10.0.0.1 (eth0) so is possible to change this address ala 'local-ip' parameter? b. As you can see all my internal hosts run private IPv4 addresses and I NAT them for both normal Internet and Internet2 which works ok but what if they want to stream something, I know Cisco could NAT multicast traffic, do you know of any solution for Linux? c. I want to export to my next router via BGP my public Internet CIDR so I have this but it doesn't work (quagga do export it correctly but problem with quagga is that it doesn't handle PIM so I have to pair it with pimd which is not as good as Xorp for tracking IGMP and it has lot less parameters to tune): policy { policy-statement to-bgp { term export { from { protocol: "connected" network4: } to { neighbor: 10.10.26.14 } then { accept } } } } bgp { export: "to-bgp" bgp-id: 10.10.26.7 local-as: 65103 peer 10.10.26.14 { local-ip: 10.10.26.7 as: 65103 next-hop: 10.10.26.7 ipv4-unicast: true ipv4-multicast: true } } I know this setup is extreme but I'm pretty sure it is a good test bed for Xorp as is a real world implementation. Thanks for the help! -otto From solca at guug.org Fri Jun 23 00:24:59 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 02:24:59 -0500 Subject: [Xorp-users] Primary interface in Linux Message-ID: <20060623072459.GB22337@guug.org> Hi again! I noted that in xorpsh when issuing commands like 'show interface address' xorp is not picking the right primary address for every interface, it appears sorted, for example I have in eth0 the next addresses: 10.0.0.12/24 10.0.0.5/24 The primary address for Linux is 10.0.0.12/24, which is correctly picked for new sockets when no source address is specified, for Xorp is the primary address is 10.0.0.5/24 which is incorrect IMO. Using latest CVS. -otto From solca at guug.org Fri Jun 23 00:17:00 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 02:17:00 -0500 Subject: [Xorp-users] rib process die Message-ID: <20060623071700.GA22337@guug.org> Hi! In latest cvs, when issuing in xorpsh the following command and quitting the pager (more) with the 'q' key the rib process die and a xorp restart is req: show route table ipv4 unicast final Just FYI I have more than 9000 routes. -otto From pavlin at icir.org Fri Jun 23 11:46:07 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 11:46:07 -0700 Subject: [Xorp-users] morekonda@gmail.com In-Reply-To: Message from Ganges of "Fri, 23 Jun 2006 11:16:11 +0530." Message-ID: <200606231846.k5NIk7Vc048933@possum.icir.org> > I'm new to the XORP community. I have tried to download the LiveCD > Rel. 1.2on Windows XP machine, it was downloaded with extenson .gz and > surprisingly > the file size was about 254MB instead of 89MB. I had tough time to burn the > CD. Later i tried the same with Linux machine and somehow got the proper > .iso image and able to burn on the CD. > > I tried to run the LiveCD image from a PC which, has latest chipsets, > motheboard and interfaces. It worked until the XOPR splash image and hung. > Appreciate any help regarding any hardware configuration/specification and > download procedure would really help me to proceed further. Look forward for > any valuable help. First, could you double-check that the MD5 checksum for the dowloaded and uncompressed image is correct: MD5 (LiveCD.iso) = 28fea148a1341ead18b2923c5bd27233 MD5 (LiveCD.iso.gz) = de2d6445f2f1e6df49543df58ad8640a The size of the above two files should be 251435008 and 93168272 respectively. There are no specific requirements about the hardware: if FreeBSD-i386 can run on it, then the LiveCD should also work. BTW, note that the XORP image is displayed while the kernel is booting so it might take up to 1-2 minutes or so until the process is completed. However, the LiveCD is based on FreeBSD-4.10 which is a little bit too old, so it could be that you have some hardware that FreeBSD-4.10 doesn't recognize. After the XORP splash image appears, can you keep pressing a number of times. If you are lucky, the image will be replaced with the kernel boot messages. Eventually, those boot messages might help finding the source of the problem. Pavlin From pavlin at icir.org Fri Jun 23 12:33:28 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 12:33:28 -0700 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 03:12:55 CDT." <20060623081255.GC22337@guug.org> Message-ID: <200606231933.k5NJXS9m051895@possum.icir.org> > I have this complex setup in Debian sarge for > my University (In Guatemala so no budget for > Cisco/Juniper routers): > > If IP address Purpose > -- ---------- ------- > eth0 10.0.0.1/24 Internal LAN > eth1 10.0.1.1/24 Internal LAN > eth2 172.16.0.1/16 Internal LAN > eth3 192.168.0.1/24 Internal LAN > eth3.2 192.168.2.1/24 Internal LAN (VLAN) > eth3.3 192.168.3.1/24 Internal LAN (VLAN) > eth3.4 192.168.4.1/24 Internal LAN (VLAN) > eth3.5 192.168.5.1/24 Internal LAN (VLAN) > eth3.6 192.168.6.1/24 Internal LAN (VLAN) > eth3.7 192.168.7.1/24 Internal LAN (VLAN) > eth3.8 192.168.8.1/24 Internal LAN (VLAN) > eth3.9 192.168.9.1/24 Internal LAN (VLAN) > eth3.10 192.168.10.1/24 Internal LAN (VLAN) > eth3.15 192.168.15.1/24 Internal LAN (VLAN) > eth4 10.0.2.1/24 Internal LAN > eth5 (unused) > eth6 10.10.26.7/28 External to Internet2 > eth7 * External to Internet > > * eth7 have 120+ assigned IPv4 address for NATing > internal hosts. > > I am using Xorp for 3 purposes: > > 1. BGP/mBGP with peer in eth6 for connecting to > Internet2. > 2. IGMP tracking for my internal hosts. > 3. PIM-SM (multicast) for my internal hosts and > for Internet2. > > BGP/mBGP works, I obtain more than 9000 routes. > IGMP and PIM-SM in my internal LANs works as > expected, I can stream to a multicast address > and any host from any internal LAN can 'see' > the stream if he join it. > > Now my problems: > > a. My next-hop router for Internet2 is Linux too > and there is no MSDP implementation for Linux > I choose to use the next-next router (Cisco) as > my RP in PIM-SM. But when sniffing the wire my > box is sending PIM-register to the Cisco (via > eth6) but with the source-address as 10.0.0.1 > (eth0) so is possible to change this address > ala 'local-ip' parameter? The address is selected automatically. Until today, the primary address of the interface toward the directly-connected source was used to originate the PIM Register messages. I just committed some changes, so now the RPF interface toward the RP is used to originate the PIM Registers. Please let me know if this doesn't solve your problem, and then we can think about adding a mechanism to manually configure that interface/address. > b. As you can see all my internal hosts run private > IPv4 addresses and I NAT them for both normal > Internet and Internet2 which works ok but what > if they want to stream something, I know Cisco > could NAT multicast traffic, do you know of any > solution for Linux? With the above modification, I believe now the PIM Registers (with the multicast data your internal hosts are originating) will be send properly to the RP. Assuming the RP doesn't filter the PIM Registers based on some rules (e.g., if there is no RPF information for the inner private source address of the multicast packets), then multicast should be working. The downside is that if the RP decides to switch to the SPT toward your sources, it can't (e.g., it might not have the RPF information for your private IP addresses), so all the multicast traffic you are originating will have the PIM Register overhead. > c. I want to export to my next router via BGP my > public Internet CIDR so I have this but it > doesn't work (quagga do export it correctly > but problem with quagga is that it doesn't > handle PIM so I have to pair it with pimd which > is not as good as Xorp for tracking IGMP and > it has lot less parameters to tune): > > policy { > policy-statement to-bgp { > term export { > from { > protocol: "connected" > network4: > } to { > neighbor: 10.10.26.14 > } then { > accept > } > } > } > } > > bgp { > export: "to-bgp" > > bgp-id: 10.10.26.7 > local-as: 65103 > > peer 10.10.26.14 { > local-ip: 10.10.26.7 > as: 65103 > next-hop: 10.10.26.7 > > ipv4-unicast: true > ipv4-multicast: true > } > } First, could you try "network4 <= " statement instead of "network4: " . If this doesn't help, for testing purpose could you remove the "to" statement and see whether it works. Pavlin > I know this setup is extreme but I'm pretty sure > it is a good test bed for Xorp as is a real world > implementation. Thanks for the help! > > -otto > > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From pavlin at icir.org Fri Jun 23 12:42:07 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 12:42:07 -0700 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: Message from Pavlin Radoslavov of "Fri, 23 Jun 2006 12:33:28 PDT." <200606231933.k5NJXS9m051895@possum.icir.org> Message-ID: <200606231942.k5NJg7WE051998@possum.icir.org> > > a. My next-hop router for Internet2 is Linux too > > and there is no MSDP implementation for Linux > > I choose to use the next-next router (Cisco) as > > my RP in PIM-SM. But when sniffing the wire my > > box is sending PIM-register to the Cisco (via > > eth6) but with the source-address as 10.0.0.1 > > (eth0) so is possible to change this address > > ala 'local-ip' parameter? > > The address is selected automatically. > Until today, the primary address of the interface toward the > directly-connected source was used to originate the PIM Register > messages. > I just committed some changes, so now the RPF interface toward the > RP is used to originate the PIM Registers. FYI, the change is in the following file and revision: Revision Changes Path 1.35 +16 -2; commitid: b5a9449c3da97ea6; xorp/pim/pim_mrt_mfc.cc Pavlin > Please let me know if this doesn't solve your problem, and then > we can think about adding a mechanism to manually configure that > interface/address. From pavlin at icir.org Fri Jun 23 13:08:59 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 13:08:59 -0700 Subject: [Xorp-users] Primary interface in Linux In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 02:24:59 CDT." <20060623072459.GB22337@guug.org> Message-ID: <200606232008.k5NK8xW9052205@possum.icir.org> > I noted that in xorpsh when issuing commands > like 'show interface address' xorp is not > picking the right primary address for every > interface, it appears sorted, for example I > have in eth0 the next addresses: > > 10.0.0.12/24 > 10.0.0.5/24 > > The primary address for Linux is 10.0.0.12/24, > which is correctly picked for new sockets when > no source address is specified, for Xorp is the > primary address is 10.0.0.5/24 which is incorrect > IMO. > > Using latest CVS. Currently there is no mechanism to explicitly specify in XORP which is the primary address of an interface. Internally, XORP chooses one of the addresses for transmission, etc (I believe the numerically smallest address is chosen). We intend to add a mechanism to manually configure the primary address of an interface. I just added a bugzilla entry about that: http://www.xorp.org/bugzilla/show_bug.cgi?id=635 Pavlin From pavlin at icir.org Fri Jun 23 13:13:55 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 13:13:55 -0700 Subject: [Xorp-users] rib process die In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 02:17:00 CDT." <20060623071700.GA22337@guug.org> Message-ID: <200606232013.k5NKDt2q052315@possum.icir.org> > In latest cvs, when issuing in xorpsh the following > command and quitting the pager (more) with the 'q' > key the rib process die and a xorp restart is req: > > show route table ipv4 unicast final > > Just FYI I have more than 9000 routes. Was there any log message from the RIB process or coredump image? Also, could you check whether you have the same issue with XORP-1.2. Recently there have been some changes to the RIB process, and there might be some recently introduced ASSERT triggering when redistributing routes (e.g., when adding the routes to the FEA or when running the program that prints them). Pavlin From solca at guug.org Fri Jun 23 14:17:13 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 16:17:13 -0500 Subject: [Xorp-users] rib process die In-Reply-To: <200606232013.k5NKDt2q052315@possum.icir.org> References: <20060623071700.GA22337@guug.org> <200606232013.k5NKDt2q052315@possum.icir.org> Message-ID: <20060623211713.GA23248@guug.org> On Fri, Jun 23, 2006 at 01:13:55PM -0700, Pavlin Radoslavov wrote: > > In latest cvs, when issuing in xorpsh the following > > command and quitting the pager (more) with the 'q' > > key the rib process die and a xorp restart is req: > > > > show route table ipv4 unicast final > > > > Just FYI I have more than 9000 routes. > > Was there any log message from the RIB process or coredump image? > Also, could you check whether you have the same issue with XORP-1.2. > > Recently there have been some changes to the RIB process, and there > might be some recently introduced ASSERT triggering when > redistributing routes (e.g., when adding the routes to the FEA or > when running the program that prints them). I redirect all output to a file but when reading it is all messed up without order, is any way to write it in sync? -otto From solca at guug.org Fri Jun 23 14:19:14 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 16:19:14 -0500 Subject: [Xorp-users] Primary interface in Linux In-Reply-To: <200606232008.k5NK8xW9052205@possum.icir.org> References: <20060623072459.GB22337@guug.org> <200606232008.k5NK8xW9052205@possum.icir.org> Message-ID: <20060623211914.GB23248@guug.org> On Fri, Jun 23, 2006 at 01:08:59PM -0700, Pavlin Radoslavov wrote: > > I noted that in xorpsh when issuing commands > > like 'show interface address' xorp is not > > picking the right primary address for every > > interface, it appears sorted, for example I > > have in eth0 the next addresses: > > > > 10.0.0.12/24 > > 10.0.0.5/24 > > > > The primary address for Linux is 10.0.0.12/24, > > which is correctly picked for new sockets when > > no source address is specified, for Xorp is the > > primary address is 10.0.0.5/24 which is incorrect > > IMO. > > > > Using latest CVS. > > Currently there is no mechanism to explicitly specify in XORP which > is the primary address of an interface. > Internally, XORP chooses one of the addresses for transmission, etc > (I believe the numerically smallest address is chosen). > > We intend to add a mechanism to manually configure the primary > address of an interface. I just added a bugzilla entry about that: > > http://www.xorp.org/bugzilla/show_bug.cgi?id=635 Excellent as is needed for complex setups. -otto From pavlin at icir.org Fri Jun 23 14:33:01 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 14:33:01 -0700 Subject: [Xorp-users] rib process die In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 16:17:13 CDT." <20060623211713.GA23248@guug.org> Message-ID: <200606232133.k5NLX1C9053049@possum.icir.org> > I redirect all output to a file but when reading it is all > messed up without order, is any way to write it in sync? Use script(1) and let the log messages appear on the terminal output: script ... exit The log messages should be saved in file "typescript". Pavlin From solca at guug.org Fri Jun 23 14:46:30 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 16:46:30 -0500 Subject: [Xorp-users] rib process die In-Reply-To: <200606232133.k5NLX1C9053049@possum.icir.org> References: <20060623211713.GA23248@guug.org> <200606232133.k5NLX1C9053049@possum.icir.org> Message-ID: <20060623214630.GC23248@guug.org> On Fri, Jun 23, 2006 at 02:33:01PM -0700, Pavlin Radoslavov wrote: > > I redirect all output to a file but when reading it is all > > messed up without order, is any way to write it in sync? > > Use script(1) and let the log messages appear on the terminal > output: > > script > > ... > > > exit > > The log messages should be saved in file "typescript". Ok, as is a very long file to attach to this mailing list I have put it here with my config.boot: http://home.galileo.edu/~solca/xorp/ Thanks. -otto From pavlin at icir.org Fri Jun 23 15:18:35 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 15:18:35 -0700 Subject: [Xorp-users] rib process die In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 16:46:30 CDT." <20060623214630.GC23248@guug.org> Message-ID: <200606232218.k5NMIZU3053560@possum.icir.org> > On Fri, Jun 23, 2006 at 02:33:01PM -0700, Pavlin Radoslavov wrote: > > > I redirect all output to a file but when reading it is all > > > messed up without order, is any way to write it in sync? > > > > Use script(1) and let the log messages appear on the terminal > > output: > > > > script > > > > ... > > > > > > exit > > > > The log messages should be saved in file "typescript". > > Ok, as is a very long file to attach to this mailing list I have > put it here with my config.boot: > > http://home.galileo.edu/~solca/xorp/ Thank you for the info. I was able to find the problematic error message from the log message. Accidentally, today we observed same error during some internal testings. This error is probably related to some of the recent CVS changes, so we will look at it. Pavlin From solca at guug.org Fri Jun 23 16:01:08 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 18:01:08 -0500 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: <200606231933.k5NJXS9m051895@possum.icir.org> References: <20060623081255.GC22337@guug.org> <200606231933.k5NJXS9m051895@possum.icir.org> Message-ID: <20060623230108.GF23248@guug.org> On Fri, Jun 23, 2006 at 12:33:28PM -0700, Pavlin Radoslavov wrote: > > c. I want to export to my next router via BGP my > > public Internet CIDR so I have this but it > > doesn't work (quagga do export it correctly > > but problem with quagga is that it doesn't > > handle PIM so I have to pair it with pimd which > > is not as good as Xorp for tracking IGMP and > > it has lot less parameters to tune): > > > > policy { > > policy-statement to-bgp { > > term export { > > from { > > protocol: "connected" > > network4: > > } to { > > neighbor: 10.10.26.14 > > } then { > > accept > > } > > } > > } > > } > > > > bgp { > > export: "to-bgp" > > > > bgp-id: 10.10.26.7 > > local-as: 65103 > > > > peer 10.10.26.14 { > > local-ip: 10.10.26.7 > > as: 65103 > > next-hop: 10.10.26.7 > > > > ipv4-unicast: true > > ipv4-multicast: true > > } > > } > > First, could you try "network4 <= " statement instead > of "network4: " . > > If this doesn't help, for testing purpose could you remove the "to" > statement and see whether it works. Using "network <= 168.234.203.0/25" statement exports every connected route. Using "network4: 168.234.203.0/25" works but I discover the problem: As you can see I have "next-hop: 10.10.26.7", the peer router said my next hop is a public IP: 168.234.203.2 so I think Xorp is ignoring the next-hop statement in BGP. Thank you. -otto From pavlin at icir.org Fri Jun 23 16:16:10 2006 From: pavlin at icir.org (Pavlin Radoslavov) Date: Fri, 23 Jun 2006 16:16:10 -0700 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 18:01:08 CDT." <20060623230108.GF23248@guug.org> Message-ID: <200606232316.k5NNGAwT054287@possum.icir.org> Otto Solares wrote: > On Fri, Jun 23, 2006 at 12:33:28PM -0700, Pavlin Radoslavov wrote: > > > c. I want to export to my next router via BGP my > > > public Internet CIDR so I have this but it > > > doesn't work (quagga do export it correctly > > > but problem with quagga is that it doesn't > > > handle PIM so I have to pair it with pimd which > > > is not as good as Xorp for tracking IGMP and > > > it has lot less parameters to tune): > > > > > > policy { > > > policy-statement to-bgp { > > > term export { > > > from { > > > protocol: "connected" > > > network4: > > > } to { > > > neighbor: 10.10.26.14 > > > } then { > > > accept > > > } > > > } > > > } > > > } > > > > > > bgp { > > > export: "to-bgp" > > > > > > bgp-id: 10.10.26.7 > > > local-as: 65103 > > > > > > peer 10.10.26.14 { > > > local-ip: 10.10.26.7 > > > as: 65103 > > > next-hop: 10.10.26.7 > > > > > > ipv4-unicast: true > > > ipv4-multicast: true > > > } > > > } > > > > First, could you try "network4 <= " statement instead > > of "network4: " . > > > > If this doesn't help, for testing purpose could you remove the "to" > > statement and see whether it works. > > Using "network <= 168.234.203.0/25" statement exports every connected > route. Using "network4: 168.234.203.0/25" works but I discover the I thought that you want to export all connected routes that fail within prefix , hence I suggested using the "<=" operator. If you want to export _only_ 168.234.203.0/25, then indeed you should use the ":" operator. > problem: As you can see I have "next-hop: 10.10.26.7", the peer > router said my next hop is a public IP: 168.234.203.2 so I think > Xorp is ignoring the next-hop statement in BGP. I am slightly confused with the above description. First, do you use any NAT between 10.10.26.7 and 10.10.26.14? >From their IP addresses I guess they are directly connected, but just want to double-check. What exactly do you mean by "the peer router said my next hop is a public IP: 168.234.203.2". Could you include the particular log message or xorpsh output that is related to this info. Thanks, Pavlin From atanu at ICSI.Berkeley.EDU Fri Jun 23 16:27:26 2006 From: atanu at ICSI.Berkeley.EDU (Atanu Ghosh) Date: Fri, 23 Jun 2006 16:27:26 -0700 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: Message from Otto Solares of "Fri, 23 Jun 2006 18:01:08 CDT." <20060623230108.GF23248@guug.org> Message-ID: <9679.1151105246@tigger.icir.org> Hi, On I-BGP peerings we don't *normally* rewrite the nexthop, I suspect that the connected route nexthop is being propagated to the peer (Maybe this is the problem with multiple address on an interface). The one case where we will rewrite the nexthop on an I-BGP peering is when the nexthop matches the peers address. In the policy statement in the then clause you could try explicitly setting the nexthop that you want. policy { policy-statement to-bgp { term export { from { protocol: "connected" network4: } to { neighbor: 10.10.26.14 } then { nexthop4: 10.10.26.7 accept } } } Atanu. >>>>> "Otto" == Otto Solares writes: Otto> On Fri, Jun 23, 2006 at 12:33:28PM -0700, Pavlin Radoslavov Otto> wrote: >> > c. I want to export to my next router via BGP my > public >> Internet CIDR so I have this but it > doesn't work (quagga do >> export it correctly > but problem with quagga is that it doesn't >> > handle PIM so I have to pair it with pimd which > is not as >> good as Xorp for tracking IGMP and > it has lot less parameters >> to tune): >> > >> > policy { > policy-statement to-bgp { > term export { > from { > >> protocol: "connected" > network4: > } to { > >> neighbor: 10.10.26.14 > } then { > accept > } > } > } > } >> > >> > bgp { > export: "to-bgp" >> > >> > bgp-id: 10.10.26.7 > local-as: 65103 >> > >> > peer 10.10.26.14 { > local-ip: 10.10.26.7 > as: 65103 > >> next-hop: 10.10.26.7 >> > >> > ipv4-unicast: true > ipv4-multicast: true > } > } >> >> First, could you try "network4 <= " statement >> instead of "network4: " . >> >> If this doesn't help, for testing purpose could you remove the >> "to" statement and see whether it works. Otto> Using "network <= 168.234.203.0/25" statement exports every Otto> connected route. Using "network4: 168.234.203.0/25" works but Otto> I discover the problem: As you can see I have "next-hop: Otto> 10.10.26.7", the peer router said my next hop is a public IP: Otto> 168.234.203.2 so I think Xorp is ignoring the next-hop Otto> statement in BGP. Otto> Thank you. Otto> -otto Otto> _______________________________________________ Xorp-users Otto> mailing list Xorp-users at xorp.org Otto> http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From solca at guug.org Fri Jun 23 16:54:37 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 18:54:37 -0500 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: <200606232316.k5NNGAwT054287@possum.icir.org> References: <20060623230108.GF23248@guug.org> <200606232316.k5NNGAwT054287@possum.icir.org> Message-ID: <20060623235437.GH23248@guug.org> On Fri, Jun 23, 2006 at 04:16:10PM -0700, Pavlin Radoslavov wrote: > Otto Solares wrote: > > > On Fri, Jun 23, 2006 at 12:33:28PM -0700, Pavlin Radoslavov wrote: > > > > c. I want to export to my next router via BGP my > > > > public Internet CIDR so I have this but it > > > > doesn't work (quagga do export it correctly > > > > but problem with quagga is that it doesn't > > > > handle PIM so I have to pair it with pimd which > > > > is not as good as Xorp for tracking IGMP and > > > > it has lot less parameters to tune): > > > > > > > > policy { > > > > policy-statement to-bgp { > > > > term export { > > > > from { > > > > protocol: "connected" > > > > network4: > > > > } to { > > > > neighbor: 10.10.26.14 > > > > } then { > > > > accept > > > > } > > > > } > > > > } > > > > } > > > > > > > > bgp { > > > > export: "to-bgp" > > > > > > > > bgp-id: 10.10.26.7 > > > > local-as: 65103 > > > > > > > > peer 10.10.26.14 { > > > > local-ip: 10.10.26.7 > > > > as: 65103 > > > > next-hop: 10.10.26.7 > > > > > > > > ipv4-unicast: true > > > > ipv4-multicast: true > > > > } > > > > } > > > > > > First, could you try "network4 <= " statement instead > > > of "network4: " . > > > > > > If this doesn't help, for testing purpose could you remove the "to" > > > statement and see whether it works. > > > > Using "network <= 168.234.203.0/25" statement exports every connected > > route. Using "network4: 168.234.203.0/25" works but I discover the > > I thought that you want to export all connected routes that fail > within prefix , hence I suggested using the "<=" > operator. If you want to export _only_ 168.234.203.0/25, then > indeed you should use the ":" operator. No, I just want to export 168.234.203.0/25 as the other connected routes are private blocks. > > problem: As you can see I have "next-hop: 10.10.26.7", the peer > > router said my next hop is a public IP: 168.234.203.2 so I think > > Xorp is ignoring the next-hop statement in BGP. > > I am slightly confused with the above description. > First, do you use any NAT between 10.10.26.7 and 10.10.26.14? > >From their IP addresses I guess they are directly connected, but > just want to double-check. Directly connected, here is an ascii pseudo-diagram: (other university) 10.10.26.3 --| (other university) 10.10.26.5 --| (other university) 10.10.26.6 --|--10.10.26.14--(Cisco) ( _my_ university) 10.10.26.7 --| (other university) 10.10.26.10 --| Every University have a public cidr block to export to 10.10.26.14. 10.10.26.14 is the route reflector for all of us. > What exactly do you mean by "the peer router said my next hop is a > public IP: 168.234.203.2". > Could you include the particular log message or xorpsh output that > is related to this info. 10.10.26.14 runs quagga, here are the logs when using Xorp in my router: BGP: 10.10.26.7 rcvd UPDATE w/ attr: nexthop 168.234.203.2, origin i, localpref 100, path BGP: 10.10.26.7 rcvd 168.234.203.0/25 BGP: 10.10.26.7 rcvd UPDATE w/ attr: , origin i, localpref 100, path BGP: 10.10.26.7 rcvd 168.234.203.0/25 As I'm in charge of 10.10.26.14 I want to switch it to Xorp too but I need to resolve my BGP problems first. -otto From solca at guug.org Fri Jun 23 16:42:40 2006 From: solca at guug.org (Otto Solares) Date: Fri, 23 Jun 2006 18:42:40 -0500 Subject: [Xorp-users] Complex BGP. mBGP and PIM-SM setup In-Reply-To: <9679.1151105246@tigger.icir.org> References: <20060623230108.GF23248@guug.org> <9679.1151105246@tigger.icir.org> Message-ID: <20060623234240.GG23248@guug.org> I set the "nexthop4: 10.10.26.7" statement in the policy but my other router reject it, looking the difference between quagga and Xorp I noted that Xorp sends an UPDATE with nexthop parameter and Quagga just send an update without parameters. Xorp sends as nexthop a public IP (the first IP in the internet iface) regardless of the mentioned statement in the policy. I forgot to tell my router is a reflector client. Thanks for the help and reply. -otto On Fri, Jun 23, 2006 at 04:27:26PM -0700, Atanu Ghosh wrote: > Hi, > > On I-BGP peerings we don't *normally* rewrite the nexthop, I suspect that > the connected route nexthop is being propagated to the peer (Maybe this > is the problem with multiple address on an interface). The one case > where we will rewrite the nexthop on an I-BGP peering is when the > nexthop matches the peers address. > > In the policy statement in the then clause you could try explicitly > setting the nexthop that you want. > > policy { > policy-statement to-bgp { > term export { > from { > protocol: "connected" > network4: > } to { > neighbor: 10.10.26.14 > } then { > nexthop4: 10.10.26.7 > accept > } > } > } > > Atanu. > > >>>>> "Otto" == Otto Solares writes: > > Otto> On Fri, Jun 23, 2006 at 12:33:28PM -0700, Pavlin Radoslavov > Otto> wrote: > >> > c. I want to export to my next router via BGP my > public > >> Internet CIDR so I have this but it > doesn't work (quagga do > >> export it correctly > but problem with quagga is that it doesn't > >> > handle PIM so I have to pair it with pimd which > is not as > >> good as Xorp for tracking IGMP and > it has lot less parameters > >> to tune): > >> > > >> > policy { > policy-statement to-bgp { > term export { > from { > > >> protocol: "connected" > network4: > } to { > > >> neighbor: 10.10.26.14 > } then { > accept > } > } > } > } > >> > > >> > bgp { > export: "to-bgp" > >> > > >> > bgp-id: 10.10.26.7 > local-as: 65103 > >> > > >> > peer 10.10.26.14 { > local-ip: 10.10.26.7 > as: 65103 > > >> next-hop: 10.10.26.7 > >> > > >> > ipv4-unicast: true > ipv4-multicast: true > } > } > >> > >> First, could you try "network4 <= " statement > >> instead of "network4: " . > >> > >> If this doesn't help, for testing purpose could you remove the > >> "to" statement and see whether it works. > > Otto> Using "network <= 168.234.203.0/25" statement exports every > Otto> connected route. Using "network4: 168.234.203.0/25" works but > Otto> I discover the problem: As you can see I have "next-hop: > Otto> 10.10.26.7", the peer router said my next hop is a public IP: > Otto> 168.234.203.2 so I think Xorp is ignoring the next-hop > Otto> statement in BGP. > > Otto> Thank you. > > Otto> -otto > > Otto> _______________________________________________ Xorp-users > Otto> mailing list Xorp-users at xorp.org > Otto> http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users From ramu.avula at gmail.com Wed Jun 28 22:15:15 2006 From: ramu.avula at gmail.com (Ramu k) Date: Thu, 29 Jun 2006 10:45:15 +0530 Subject: [Xorp-users] No buffer space available Message-ID: Hi , I am running XORP for past two weeks with one outgoing interface. Now i have configured the XORP with 10 VLANS and it is not scaling and coming out with the following error. ERROR xorp_fea:4373 MFEA +714 mfea_proto_comm.cc join_multicast_group ] Cannot join group 224.0.0.1 on vif eth0.11: No buffer space available is it a memory problem(i have 1GB RAM) or thing wrong with the xorp. BTW how many interfaces does XORP supports and is there any limit on no of interfaces. with 5 VLANS its working fine. Thanks for any Help in this regard. Ramu. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.ICSI.Berkeley.EDU/pipermail/xorp-users/attachments/20060629/9ae48ccb/attachment.html From solca at guug.org Wed Jun 28 22:32:05 2006 From: solca at guug.org (Otto Solares) Date: Thu, 29 Jun 2006 00:32:05 -0500 Subject: [Xorp-users] No buffer space available In-Reply-To: References: Message-ID: <20060629053205.GA2730@guug.org> I think that problem is mentioned in the ERRATA file, something to do with the multicast membership limit. -otto On Thu, Jun 29, 2006 at 10:45:15AM +0530, Ramu k wrote: > Hi , > > I am running XORP for past two weeks with one outgoing interface. > Now i have configured the XORP with 10 VLANS and it is not scaling and > coming out with the following error. > > ERROR xorp_fea:4373 MFEA +714 mfea_proto_comm.cc join_multicast_group ] > Cannot join group 224.0.0.1 > on vif eth0.11: No buffer space available > > is it a memory problem(i have 1GB RAM) or thing wrong with the xorp. > BTW how many interfaces does XORP supports and is there any limit on > no of interfaces. with 5 VLANS its working fine. > > > Thanks for any Help in this regard. > Ramu. > _______________________________________________ > Xorp-users mailing list > Xorp-users at xorp.org > http://mailman.ICSI.Berkeley.EDU/mailman/listinfo/xorp-users