From cammerra at gmail.com Wed May 4 01:07:55 2011 From: cammerra at gmail.com (Antonie falco) Date: Wed, 4 May 2011 15:07:55 +1000 Subject: [rbak-nsp] Compact Flash Card Speeds on XCRP3 Controller (PIO, UDMA, WDMA) Message-ID: Hello All, Can anyone confirm if the Redback XCRP3 utilises PIO (Mode 6), UDMA (Mode 5) or (WDMA Mode 3,4) for the Host to Card communication? Asuming support for all PIO,WDMA & UMDA can anyone confirm if a slower CF which used PIO Mode 6 for VxWorks would cause any degridation in the performance of a of the overall controller whilst up apart from the writing to the CF for configuration changes or reading of the configuration on boot? I believe the CF is only used on boot to load the config into RAM and for the writing and reading of configuration changes. Regards, Antonie Falco -------------- next part -------------- An HTML attachment was scrubbed... URL: From soltanian at gmail.com Tue May 10 10:12:53 2011 From: soltanian at gmail.com (Alireza Soltanian) Date: Tue, 10 May 2011 18:42:53 +0430 Subject: [rbak-nsp] PWFQ on 10GE card SE800 Message-ID: <002a01cc0f1c$5d5d5380$1817fa80$@gmail.com> Hi We have a 10GE-1 card on SE-800, We want to configure Hierarchical qos on this port, so we need to run PWFQ. When we want to configure it, we receive an error about not supporting this feature. Anybody has any solution? -------------- next part -------------- An HTML attachment was scrubbed... URL: From soltanian at gmail.com Tue May 10 10:35:43 2011 From: soltanian at gmail.com (Alireza Soltanian) Date: Tue, 10 May 2011 19:05:43 +0430 Subject: [rbak-nsp] PWFQ on 10GE card SE800 In-Reply-To: References: <002a01cc0f1c$5d5d5380$1817fa80$@gmail.com> Message-ID: <004101cc0f1f$8d288eb0$a779ac10$@gmail.com> Thank you Frans, So briefly I must have 10G-4 instead of 10G-1. I also want to ask, is there any 10G-4 for SE-800? because based on my documents, this card is not supported on SE-800, It is only supported on SE-600 and SE-1200 Regards Ali From: Frans Legdeur [mailto:frans at falco-networks.com] Sent: Tuesday, May 10, 2011 6:55 PM To: Alireza Soltanian; redback-nsp at puck.nether.net Subject: Re: [rbak-nsp] PWFQ on 10GE card SE800 Hi Alieza, Unfortunately, both single port 10G cards use PPA-2, which doesn't support your requested PWFQ. You would need the 4 port 10G card: 10ge-4-port with PPA-3. Per manual: The 1-port 10 Gigabit Ethernet (1x10GE) card is designed for traffic management using the second-generation PPAs. This card has an increased minimum memory capacity of 1 GB and can process data internally to match the speed of the port, which runs at 10 Gbps. The 4-port 10GE is a DDR-based card designed to address the requirements for applications and services with large volumes of network traffic. Each 4-port 10GE DDR card is optimized to deliver Ethernet services, operates as an uplink to other switches and routers in the network or for interconnectivity between two SmartEdge routers. This PPA3-based card has an increased port density compared with previous 10GE cards. It supports up to 48K circuits per card. This 4-port 10GE DDR card is capable of a maximum throughput of 20 Gbps. It can be used in ingress oversubscription mode at a maximum of 2 to 1 (40 Gbps over four 10GE faceplate ports; 20 Gbps within the card and across the backplane). Ports on this card can be configured in one of two ingress oversubscription modes for incoming traffic: priority-based packet discard for traffic from trusted networks and random-based packet discard from unfiltered nontrusted networks. The 1-port 10GE/OC-192c DDR-based card designed for traffic management using second-generation PPAs. This multimode DDR card supports the 10GE LAN-PHY, 10GE WAN-PHY, 10GE-DWDM, POS OC-192c, OC-192c DWDM, or OTN-DWDM modes for the SmartEdge routers. This card supports a minimum of 1 GB of memory capacity and can process data internally to match the speed of the port - 10.3125 Gbps in 10GE LAN-PHY or 10GE-DWDM mode; 9.953 Gbps in 10GE WAN-PHY, POS OC-192c, or OC-192c DWDM mode; and 11.0957 Gbps in OTN-DWDM mode. For Ethernet LAN-PHY and WAN-PHY modes, this card supports dot1q, PPPoE, and plain Ethernet encapsulations. For POS mode, it supports PPP, HDLC, and FR encapsulations. Kind regards, Frans. _____ From: Alireza Soltanian Date: Tue, 10 May 2011 18:42:53 +0430 To: Subject: [rbak-nsp] PWFQ on 10GE card SE800 Hi We have a 10GE-1 card on SE-800, We want to configure Hierarchical qos on this port, so we need to run PWFQ. When we want to configure it, we receive an error about not supporting this feature. Anybody has any solution? _____ _______________________________________________ redback-nsp mailing list redback-nsp at puck.nether.net https://puck.nether.net/mailman/listinfo/redback-nsp -------------- next part -------------- An HTML attachment was scrubbed... URL: From frans at falco-networks.com Tue May 10 10:25:05 2011 From: frans at falco-networks.com (Frans Legdeur) Date: Tue, 10 May 2011 16:25:05 +0200 Subject: [rbak-nsp] PWFQ on 10GE card SE800 In-Reply-To: <002a01cc0f1c$5d5d5380$1817fa80$@gmail.com> Message-ID: Hi Alieza, Unfortunately, both single port 10G cards use PPA-2, which doesn?t support your requested PWFQ. You would need the 4 port 10G card: 10ge-4-port with PPA-3. Per manual: The 1-port 10 Gigabit Ethernet (1x10GE) card is designed for traffic management using the second-generation PPAs. This card has an increased minimum memory capacity of 1 GB and can process data internally to match the speed of the port, which runs at 10 Gbps. The 4-port 10GE is a DDR-based card designed to address the requirements for applications and services with large volumes of network traffic. Each 4-port 10GE DDR card is optimized to deliver Ethernet services, operates as an uplink to other switches and routers in the network or for interconnectivity between two SmartEdge routers. This PPA3-based card has an increased port density compared with previous 10GE cards. It supports up to 48K circuits per card. This 4-port 10GE DDR card is capable of a maximum throughput of 20 Gbps. It can be used in ingress oversubscription mode at a maximum of 2 to 1 (40 Gbps over four 10GE faceplate ports; 20 Gbps within the card and across the backplane). Ports on this card can be configured in one of two ingress oversubscription modes for incoming traffic: priority-based packet discard for traffic from trusted networks and random-based packet discard from unfiltered nontrusted networks. The 1-port 10GE/OC-192c DDR-based card designed for traffic management using second-generation PPAs. This multimode DDR card supports the 10GE LAN-PHY, 10GE WAN-PHY, 10GE-DWDM, POS OC-192c, OC-192c DWDM, or OTN-DWDM modes for the SmartEdge routers. This card supports a minimum of 1 GB of memory capacity and can process data internally to match the speed of the port ? 10.3125 Gbps in 10GE LAN-PHY or 10GE-DWDM mode; 9.953 Gbps in 10GE WAN-PHY, POS OC-192c, or OC-192c DWDM mode; and 11.0957 Gbps in OTN-DWDM mode. For Ethernet LAN-PHY and WAN-PHY modes, this card supports dot1q, PPPoE, and plain Ethernet encapsulations. For POS mode, it supports PPP, HDLC, and FR encapsulations. Kind regards, Frans. From: Alireza Soltanian Date: Tue, 10 May 2011 18:42:53 +0430 To: Subject: [rbak-nsp] PWFQ on 10GE card SE800 Hi We have a 10GE-1 card on SE-800, We want to configure Hierarchical qos on this port, so we need to run PWFQ. When we want to configure it, we receive an error about not supporting this feature. Anybody has any solution? _______________________________________________ redback-nsp mailing list redback-nsp at puck.nether.net https://puck.nether.net/mailman/listinfo/redback-nsp -------------- next part -------------- An HTML attachment was scrubbed... URL: From frans at falco-networks.com Tue May 10 10:51:41 2011 From: frans at falco-networks.com (Frans Legdeur) Date: Tue, 10 May 2011 16:51:41 +0200 Subject: [rbak-nsp] PWFQ on 10GE card SE800 In-Reply-To: <004101cc0f1f$8d288eb0$a779ac10$@gmail.com> Message-ID: True, Although I don?t see why these cards should not work in an SE-800 chassis. All cards interconnect to each other and negotiate the communication speed over the backplane with each other card individually. I can imagine that if you have two of these cards in the SE-800, that they fail to interconnect at 20G, which they are designed for. But if you use a single 4 port 10G card in the SE-800, combined with other cards, like single port 10G / 1 G cards, it shouldn?t be any problem. Kind regards, Frans. From: Alireza Soltanian Date: Tue, 10 May 2011 19:05:43 +0430 To: Frans Legdeur , Subject: RE: [rbak-nsp] PWFQ on 10GE card SE800 Thank you Frans, So briefly I must have 10G-4 instead of 10G-1. I also want to ask, is there any 10G-4 for SE-800? because based on my documents, this card is not supported on SE-800, It is only supported on SE-600 and SE-1200 Regards Ali From: Frans Legdeur [mailto:frans at falco-networks.com] Sent: Tuesday, May 10, 2011 6:55 PM To: Alireza Soltanian; redback-nsp at puck.nether.net Subject: Re: [rbak-nsp] PWFQ on 10GE card SE800 Hi Alieza, Unfortunately, both single port 10G cards use PPA-2, which doesn?t support your requested PWFQ. You would need the 4 port 10G card: 10ge-4-port with PPA-3. Per manual: The 1-port 10 Gigabit Ethernet (1x10GE) card is designed for traffic management using the second-generation PPAs. This card has an increased minimum memory capacity of 1 GB and can process data internally to match the speed of the port, which runs at 10 Gbps. The 4-port 10GE is a DDR-based card designed to address the requirements for applications and services with large volumes of network traffic. Each 4-port 10GE DDR card is optimized to deliver Ethernet services, operates as an uplink to other switches and routers in the network or for interconnectivity between two SmartEdge routers. This PPA3-based card has an increased port density compared with previous 10GE cards. It supports up to 48K circuits per card. This 4-port 10GE DDR card is capable of a maximum throughput of 20 Gbps. It can be used in ingress oversubscription mode at a maximum of 2 to 1 (40 Gbps over four 10GE faceplate ports; 20 Gbps within the card and across the backplane). Ports on this card can be configured in one of two ingress oversubscription modes for incoming traffic: priority-based packet discard for traffic from trusted networks and random-based packet discard from unfiltered nontrusted networks. The 1-port 10GE/OC-192c DDR-based card designed for traffic management using second-generation PPAs. This multimode DDR card supports the 10GE LAN-PHY, 10GE WAN-PHY, 10GE-DWDM, POS OC-192c, OC-192c DWDM, or OTN-DWDM modes for the SmartEdge routers. This card supports a minimum of 1 GB of memory capacity and can process data internally to match the speed of the port ? 10.3125 Gbps in 10GE LAN-PHY or 10GE-DWDM mode; 9.953 Gbps in 10GE WAN-PHY, POS OC-192c, or OC-192c DWDM mode; and 11.0957 Gbps in OTN-DWDM mode. For Ethernet LAN-PHY and WAN-PHY modes, this card supports dot1q, PPPoE, and plain Ethernet encapsulations. For POS mode, it supports PPP, HDLC, and FR encapsulations. Kind regards, Frans. From: Alireza Soltanian Date: Tue, 10 May 2011 18:42:53 +0430 To: Subject: [rbak-nsp] PWFQ on 10GE card SE800 Hi We have a 10GE-1 card on SE-800, We want to configure Hierarchical qos on this port, so we need to run PWFQ. When we want to configure it, we receive an error about not supporting this feature. Anybody has any solution? _______________________________________________ redback-nsp mailing list redback-nsp at puck.nether.net https://puck.nether.net/mailman/listinfo/redback-nsp -------------- next part -------------- An HTML attachment was scrubbed... URL: From prapti.soe at gmail.com Wed May 11 00:27:43 2011 From: prapti.soe at gmail.com (soe prapti) Date: Wed, 11 May 2011 11:27:43 +0700 Subject: [rbak-nsp] NMI Interrupt Message-ID: Hi, I have case in SE1200, the case begin at 26 of March 2011, start from XCRP fail over then port 10G - 4 port was up and down simultanously . In the show system nvlog i find like this : Mar 12 21:18:27 NMI interrupt: in kernel mode Mar 12 21:18:27 status=0x4c800a5, cause=0x20008c08 Mar 12 21:18:27 epc=0xa800000080297960 Mar 12 21:18:27 pid=746 cmd=syslogdMar 12 21:18:27 usp=0xffffffd430 Mar 12 21:18:27 ksp=0xc00000000e43b998 Mar 12 21:21:28 *** TOMTE WDOG RESET : booting up SEOS-6.2.1.2-Release Mar 12 21:21:29 DISK INFO: wd2: <00001011540001139129> <1002092P> Mar 12 21:21:29 DISK INFO: wd1: <00001011510102159988> <1002092P> Mar 16 15:25:43 NMI interrupt: in kernel mode Mar 16 15:25:43 status=0x4c800a4, cause=0x8c00 Mar 16 15:25:43 epc=0xa800000080296d70 Mar 16 15:25:43 curproc == NULL Mar 16 15:25:43 curlwp == NULL Mar 16 15:25:43 ksp=0xa80000009c98ba40 Mar 16 15:28:44 *** TOMTE WDOG RESET : booting up SEOS-6.2.1.2-Release Mar 16 15:28:45 DISK INFO: wd2: <00001011540001139129> <1002092P> Mar 16 15:28:45 DISK INFO: wd1: <00001011510102159988> <1002092P> Mar 21 23:30:31 NMI interrupt: in kernel mode Mar 21 23:30:31 status=0x4c800a4, cause=0x20008d08 Mar 21 23:30:31 epc=0xa800000080296d94 Mar 21 23:30:31 pid=1365 cmd=igmpdMar 21 23:30:31 usp=0x1006fd4e0 Mar 21 23:30:31 ksp=0xc00000000e5dfa20 Mar 21 23:30:31 xc4: dumpsys called Mar 21 23:30:31 dumping to dev 4,33 offset 8 Can somebody help me about information of this log ? and how to solve it -- Regards, Suparti +62 81384850602 -- Regards, Suparti 081384850602 -------------- next part -------------- An HTML attachment was scrubbed... URL: From oleg at slavkin.info Wed May 11 01:21:51 2011 From: oleg at slavkin.info (=?KOI8-R?B?88zB18vJziDvzMXH?=) Date: Wed, 11 May 2011 09:21:51 +0400 Subject: [rbak-nsp] NMI Interrupt In-Reply-To: References: Message-ID: Please provide printout "show tech" //br Slavkin Oleg 2011/5/11 soe prapti > Hi, > > > I have case in SE1200, the case begin at 26 of March 2011, start from XCRP > fail over then port 10G - 4 port was up and down simultanously . In the show > system nvlog i find like this : > > Mar 12 21:18:27 NMI interrupt: in kernel mode > Mar 12 21:18:27 status=0x4c800a5, cause=0x20008c08 > Mar 12 21:18:27 epc=0xa800000080297960 > Mar 12 21:18:27 pid=746 cmd=syslogdMar 12 21:18:27 usp=0xffffffd430 Mar > 12 21:18:27 ksp=0xc00000000e43b998 > Mar 12 21:21:28 *** TOMTE WDOG RESET : booting up SEOS-6.2.1.2-Release > Mar 12 21:21:29 DISK INFO: wd2: <00001011540001139129> > <1002092P> > Mar 12 21:21:29 DISK INFO: wd1: <00001011510102159988> > <1002092P> > Mar 16 15:25:43 NMI interrupt: in kernel mode > Mar 16 15:25:43 status=0x4c800a4, cause=0x8c00 > Mar 16 15:25:43 epc=0xa800000080296d70 > Mar 16 15:25:43 curproc == NULL Mar 16 15:25:43 curlwp == NULL Mar 16 > 15:25:43 ksp=0xa80000009c98ba40 > Mar 16 15:28:44 *** TOMTE WDOG RESET : booting up SEOS-6.2.1.2-Release > Mar 16 15:28:45 DISK INFO: wd2: <00001011540001139129> > <1002092P> > Mar 16 15:28:45 DISK INFO: wd1: <00001011510102159988> > <1002092P> > Mar 21 23:30:31 NMI interrupt: in kernel mode > Mar 21 23:30:31 status=0x4c800a4, cause=0x20008d08 > Mar 21 23:30:31 epc=0xa800000080296d94 > Mar 21 23:30:31 pid=1365 cmd=igmpdMar 21 23:30:31 usp=0x1006fd4e0 Mar > 21 23:30:31 ksp=0xc00000000e5dfa20 > Mar 21 23:30:31 xc4: dumpsys called > Mar 21 23:30:31 > dumping to dev 4,33 offset 8 > > > Can somebody help me about information of this log ? and how to solve it > > > -- > Regards, > > Suparti > +62 81384850602 > > -- > Regards, > > Suparti > 081384850602 > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From marcin at leon.pl Wed May 11 15:14:21 2011 From: marcin at leon.pl (Marcin Kuczera) Date: Wed, 11 May 2011 21:14:21 +0200 Subject: [rbak-nsp] PWFQ on 10GE card SE800 In-Reply-To: References: Message-ID: <4DCAE00D.9020509@leon.pl> Frans Legdeur wrote: > True, > > Although I don?t see why these cards should not work in an SE-800 chassis. > All cards interconnect to each other and negotiate the communication > speed over the backplane with each other card individually. > I can imagine that if you have two of these cards in the SE-800, that > they fail to interconnect at 20G, which they are designed for. > But if you use a single 4 port 10G card in the SE-800, combined with > other cards, like single port 10G / 1 G cards, it shouldn?t be any problem. That could be refused by software. However, 20G cards should have also 10G backplane interfaces.. just marketing I suppose. But, what you can do is to buy just SE600 or SE1200 chassis and move all cards (including XCRP3/4) to new chassis. This will be the cheapest way to use 4x10G cards I suppose. Regards, Marcin > > > Kind regards, > > > Frans. > > > ------------------------------------------------------------------------ > *From: *Alireza Soltanian > *Date: *Tue, 10 May 2011 19:05:43 +0430 > *To: *Frans Legdeur , > > *Subject: *RE: [rbak-nsp] PWFQ on 10GE card SE800 > > Thank you Frans, > > So briefly I must have 10G-4 instead of 10G-1. I also want to ask, is > there any 10G-4 for SE-800? > because based on my documents, this card is not supported on SE-800, It > is only supported on SE-600 and SE-1200 > > Regards > Ali > > > > > *From:* Frans Legdeur [mailto:frans at falco-networks.com] > *Sent:* Tuesday, May 10, 2011 6:55 PM > *To:* Alireza Soltanian; redback-nsp at puck.nether.net > *Subject:* Re: [rbak-nsp] PWFQ on 10GE card SE800 > > Hi Alieza, > > Unfortunately, both single port 10G cards use PPA-2, which doesn?t > support your requested PWFQ. > You would need the 4 port 10G card: 10ge-4-port with PPA-3. > > Per manual: > The 1-port 10 Gigabit Ethernet (1x10GE) card is designed for traffic > management using the second-generation PPAs. This card has an increased > minimum memory capacity of 1 GB and can process data internally to match > the speed of the port, which runs at 10 Gbps. > > The 4-port 10GE is a DDR-based card designed to address the requirements > for applications and services with large volumes of network traffic. > Each 4-port 10GE DDR card is optimized to deliver Ethernet services, > operates as an uplink to other switches and routers in the network or > for interconnectivity between two SmartEdge routers. > This PPA3-based card has an increased port density compared with > previous 10GE cards. It supports up to 48K circuits per card. This > 4-port 10GE DDR card is capable of a maximum throughput of 20 Gbps. It > can be used in ingress oversubscription mode at a maximum of 2 to 1 (40 > Gbps over four 10GE faceplate ports; 20 Gbps within the card and across > the backplane). Ports on this card can be configured in one of two > ingress oversubscription modes for incoming traffic: priority-based > packet discard for traffic from trusted networks and random-based packet > discard from unfiltered nontrusted networks. > > The 1-port 10GE/OC-192c DDR-based card designed for traffic management > using second-generation PPAs. This multimode DDR card supports the 10GE > LAN-PHY, 10GE WAN-PHY, 10GE-DWDM, POS OC-192c, OC-192c DWDM, or OTN-DWDM > modes for the SmartEdge routers. > > This card supports a minimum of 1 GB of memory capacity and can process > data internally to match the speed of the port ? 10.3125 Gbps in 10GE > LAN-PHY or 10GE-DWDM mode; 9.953 Gbps in 10GE WAN-PHY, POS OC-192c, or > OC-192c DWDM mode; and 11.0957 Gbps in OTN-DWDM mode. > > For Ethernet LAN-PHY and WAN-PHY modes, this card supports dot1q, PPPoE, > and plain Ethernet encapsulations. For POS mode, it supports PPP, HDLC, > and FR encapsulations. > > Kind regards, > > Frans. > > ------------------------------------------------------------------------ > > *From: *Alireza Soltanian > *Date: *Tue, 10 May 2011 18:42:53 +0430 > *To: * > *Subject: *[rbak-nsp] PWFQ on 10GE card SE800 > > Hi > > We have a 10GE-1 card on SE-800, We want to configure Hierarchical qos > on this port, so we need to run PWFQ. When we want to configure it, we > receive an error about not supporting this feature. > > > > Anybody has any solution? > > > ------------------------------------------------------------------------ > > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > > > ------------------------------------------------------------------------ > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp From rob at wonk.org Wed May 11 20:50:03 2011 From: rob at wonk.org (Rob Wise) Date: Thu, 12 May 2011 10:50:03 +1000 Subject: [rbak-nsp] PWFQ on 10GE card SE800 In-Reply-To: <4DCAE00D.9020509@leon.pl> References: <4DCAE00D.9020509@leon.pl> Message-ID: Hi all, The newest cards such as 4x10G will not work in the older SE400/800 chassis. Also some of the oldest cards will not work in the newer SE600/1200 chassis. This is due to physical differences with the backplanes. The middle generation of cards such as the 10x1G will work in all chassis. It's best to contact Ericsson support or your account manager to confirm compatibility for your particular combination of cards and chassis. Cheers, Rob On Thu, May 12, 2011 at 5:14 AM, Marcin Kuczera wrote: > Frans Legdeur wrote: > >> True, >> >> Although I don?t see why these cards should not work in an SE-800 chassis. >> All cards interconnect to each other and negotiate the communication speed >> over the backplane with each other card individually. >> I can imagine that if you have two of these cards in the SE-800, that they >> fail to interconnect at 20G, which they are designed for. >> But if you use a single 4 port 10G card in the SE-800, combined with other >> cards, like single port 10G / 1 G cards, it shouldn?t be any problem. >> > > That could be refused by software. > However, 20G cards should have also 10G backplane interfaces.. just > marketing I suppose. > > But, what you can do is to buy just SE600 or SE1200 chassis and move all > cards (including XCRP3/4) to new chassis. > This will be the cheapest way to use 4x10G cards I suppose. > > Regards, > Marcin > > >> >> Kind regards, >> >> >> Frans. >> >> >> ------------------------------------------------------------------------ >> *From: *Alireza Soltanian >> *Date: *Tue, 10 May 2011 19:05:43 +0430 >> *To: *Frans Legdeur , < >> redback-nsp at puck.nether.net> >> *Subject: *RE: [rbak-nsp] PWFQ on 10GE card SE800 >> >> Thank you Frans, >> So briefly I must have 10G-4 instead of 10G-1. I also want to ask, is >> there any 10G-4 for SE-800? >> because based on my documents, this card is not supported on SE-800, It >> is only supported on SE-600 and SE-1200 >> Regards >> Ali >> >> *From:* Frans Legdeur [mailto:frans at falco-networks.com] >> *Sent:* Tuesday, May 10, 2011 6:55 PM >> *To:* Alireza Soltanian; redback-nsp at puck.nether.net >> *Subject:* Re: [rbak-nsp] PWFQ on 10GE card SE800 >> >> Hi Alieza, >> >> Unfortunately, both single port 10G cards use PPA-2, which doesn?t support >> your requested PWFQ. >> You would need the 4 port 10G card: 10ge-4-port with PPA-3. >> >> Per manual: >> The 1-port 10 Gigabit Ethernet (1x10GE) card is designed for traffic >> management using the second-generation PPAs. This card has an increased >> minimum memory capacity of 1 GB and can process data internally to match the >> speed of the port, which runs at 10 Gbps. >> >> The 4-port 10GE is a DDR-based card designed to address the requirements >> for applications and services with large volumes of network traffic. Each >> 4-port 10GE DDR card is optimized to deliver Ethernet services, operates as >> an uplink to other switches and routers in the network or for >> interconnectivity between two SmartEdge routers. >> This PPA3-based card has an increased port density compared with previous >> 10GE cards. It supports up to 48K circuits per card. This 4-port 10GE DDR >> card is capable of a maximum throughput of 20 Gbps. It can be used in >> ingress oversubscription mode at a maximum of 2 to 1 (40 Gbps over four 10GE >> faceplate ports; 20 Gbps within the card and across the backplane). Ports on >> this card can be configured in one of two ingress oversubscription modes for >> incoming traffic: priority-based packet discard for traffic from trusted >> networks and random-based packet discard from unfiltered nontrusted >> networks. >> >> The 1-port 10GE/OC-192c DDR-based card designed for traffic management >> using second-generation PPAs. This multimode DDR card supports the 10GE >> LAN-PHY, 10GE WAN-PHY, 10GE-DWDM, POS OC-192c, OC-192c DWDM, or OTN-DWDM >> modes for the SmartEdge routers. >> >> This card supports a minimum of 1 GB of memory capacity and can process >> data internally to match the speed of the port ? 10.3125 Gbps in 10GE >> LAN-PHY or 10GE-DWDM mode; 9.953 Gbps in 10GE WAN-PHY, POS OC-192c, or >> OC-192c DWDM mode; and 11.0957 Gbps in OTN-DWDM mode. >> >> For Ethernet LAN-PHY and WAN-PHY modes, this card supports dot1q, PPPoE, >> and plain Ethernet encapsulations. For POS mode, it supports PPP, HDLC, and >> FR encapsulations. >> >> Kind regards, >> >> Frans. >> >> ------------------------------------------------------------------------ >> >> *From: *Alireza Soltanian >> *Date: *Tue, 10 May 2011 18:42:53 +0430 >> *To: * >> *Subject: *[rbak-nsp] PWFQ on 10GE card SE800 >> >> Hi >> We have a 10GE-1 card on SE-800, We want to configure Hierarchical qos on >> this port, so we need to run PWFQ. When we want to configure it, we receive >> an error about not supporting this feature. >> Anybody has any solution? >> >> ------------------------------------------------------------------------ >> >> >> _______________________________________________ >> redback-nsp mailing list >> redback-nsp at puck.nether.net >> https://puck.nether.net/mailman/listinfo/redback-nsp >> >> >> ------------------------------------------------------------------------ >> >> >> _______________________________________________ >> redback-nsp mailing list >> redback-nsp at puck.nether.net >> https://puck.nether.net/mailman/listinfo/redback-nsp >> > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > -------------- next part -------------- An HTML attachment was scrubbed... URL: From delivery at yousendit.com Fri May 13 01:21:44 2011 From: delivery at yousendit.com (prapti ST) Date: Thu, 12 May 2011 22:21:44 -0700 Subject: [rbak-nsp] NMI Interruptable Message-ID: <257994303.1305264104674.JavaMail.ysi@sjcprdngw01.prod.yousendit.com> Dear Slavkin Oleg, Kindly please find attachment of technical support Thank you very much You have 1 file waiting for download: tech support.rar You can copy and paste the following link in your browser to retrieve your file. The link will expire on May 19, 2011 22:21 PDT. https://rcpt.yousendit.com/1120507489/6d481a12e0d6473baedfc1b45156560a YouSendIt, Inc.: http://www.yousendit.com Privacy Policy : http://www.yousendit.com/aboutus/legal/privacy 1919 S. Bascom Ave., 3rd Floor, Campbell, CA 95008 -------------- next part -------------- An HTML attachment was scrubbed... URL: From prapti.soe at gmail.com Fri May 13 01:22:09 2011 From: prapti.soe at gmail.com (soe prapti) Date: Fri, 13 May 2011 12:22:09 +0700 Subject: [rbak-nsp] NMI Interrupt In-Reply-To: References: Message-ID: Dear Slavkin Oleg, Kindly please find attachment of technical support on https://www.yousendit.com/download/MEtRZHluT2I1R05jR0E9PQ Thank you very much On Wed, May 11, 2011 at 12:21 PM, ??????? ???? wrote: > Please provide printout "show tech" > //br > Slavkin Oleg > > > 2011/5/11 soe prapti > >> Hi, >> >> >> I have case in SE1200, the case begin at 26 of March 2011, start from >> XCRP fail over then port 10G - 4 port was up and down simultanously . In the >> show system nvlog i find like this : >> >> Mar 12 21:18:27 NMI interrupt: in kernel mode >> Mar 12 21:18:27 status=0x4c800a5, cause=0x20008c08 >> Mar 12 21:18:27 epc=0xa800000080297960 >> Mar 12 21:18:27 pid=746 cmd=syslogdMar 12 21:18:27 usp=0xffffffd430 >> Mar 12 21:18:27 ksp=0xc00000000e43b998 >> Mar 12 21:21:28 *** TOMTE WDOG RESET : booting up SEOS-6.2.1.2-Release >> Mar 12 21:21:29 DISK INFO: wd2: <00001011540001139129> >> <1002092P> >> Mar 12 21:21:29 DISK INFO: wd1: <00001011510102159988> >> <1002092P> >> Mar 16 15:25:43 NMI interrupt: in kernel mode >> Mar 16 15:25:43 status=0x4c800a4, cause=0x8c00 >> Mar 16 15:25:43 epc=0xa800000080296d70 >> Mar 16 15:25:43 curproc == NULL Mar 16 15:25:43 curlwp == NULL Mar 16 >> 15:25:43 ksp=0xa80000009c98ba40 >> Mar 16 15:28:44 *** TOMTE WDOG RESET : booting up SEOS-6.2.1.2-Release >> Mar 16 15:28:45 DISK INFO: wd2: <00001011540001139129> >> <1002092P> >> Mar 16 15:28:45 DISK INFO: wd1: <00001011510102159988> >> <1002092P> >> Mar 21 23:30:31 NMI interrupt: in kernel mode >> Mar 21 23:30:31 status=0x4c800a4, cause=0x20008d08 >> Mar 21 23:30:31 epc=0xa800000080296d94 >> Mar 21 23:30:31 pid=1365 cmd=igmpdMar 21 23:30:31 usp=0x1006fd4e0 Mar >> 21 23:30:31 ksp=0xc00000000e5dfa20 >> Mar 21 23:30:31 xc4: dumpsys called >> Mar 21 23:30:31 >> dumping to dev 4,33 offset 8 >> >> >> Can somebody help me about information of this log ? and how to solve it >> >> >> -- >> Regards, >> >> Suparti >> +62 81384850602 >> >> -- >> Regards, >> >> Suparti >> 081384850602 >> >> _______________________________________________ >> redback-nsp mailing list >> redback-nsp at puck.nether.net >> https://puck.nether.net/mailman/listinfo/redback-nsp >> >> > -- Regards, Suparti 081384850602 -------------- next part -------------- An HTML attachment was scrubbed... URL: From mgage at localtel.net Fri May 13 10:54:56 2011 From: mgage at localtel.net (Michael J. Gage) Date: Fri, 13 May 2011 14:54:56 +0000 Subject: [rbak-nsp] Link-Group Questions Message-ID: I have two Redback SE800 routers, and I am using a 4 port link-group between them. Edge Router (hostname: Itt - SEOS-6.1.3.6-Release) using 2x XCRP3 & 4x ge-4-port. Subscriber Router (hostname: Thing - - SEOS-6.1.5.7p5-Release) using 2x XCRP4 & 1x ge-10-port. This issue is load-balancing across all 4 links. One direction I balance across all 4 and the other it only uses 2. Edge --> Subscriber: 1. 1/3 (Up )| 23.71% 2. 2/3 (Up )| 19.89% 3. 3/3 (Up )| 0.00% 4. 4/3 (Up )| 0.00% Subscriber --> Edge: 1. 1/7 (Up )| 2.60% 2. 1/8 (Up )| 1.85% 3. 1/9 (Up )| 2.75% 4. 1/10 (Up )| 2.25% Any thoughts? Michael J. Gage Network Operations Manger LocalTel Communications (509) 888-9208 - Office (509) 669-8482 - Mobile Current configuration: (Edge Side) ! link-group Thing dot1q dot1q pvc 2 bind interface Core bgp ! dot1q pvc 3 bind interface CCFiber bgp ! dot1q pvc 4 bind interface DCFiber bgp ! dot1q pvc 5 bind interface GCFiber bgp ! dot1q pvc 6 bind interface copper bgp ! dot1q pvc 13 bind interface ccfiber noauth ! dot1q pvc 14 bind interface dcfiber noauth ! dot1q pvc 15 bind interface gcfiber noauth ! maximum-links 4 lacp active ! card ge-4-port 1 port ethernet 1/3 link-group Thing ! card ge-4-port 2 port ethernet 2/3 link-group Thing ! card ge-4-port 3 port ethernet 3/3 link-group Thing ! card ge-4-port 4 port ethernet 4/3 link-group Thing ! Current configuration: (Subscriber Side) ! link-group Itt dot1q dot1q pvc 3 bind interface Gateway ccfiber dot1q pvc 4 bind interface Gateway dcfiber dot1q pvc 5 bind interface Gateway gcfiber dot1q pvc 6 bind interface Gateway copper dot1q pvc 13 bind interface NoAuth ccfiber dot1q pvc 14 bind interface NoAuth dcfiber dot1q pvc 15 bind interface NoAuth gcfiber maximum-links 4 ! card ge-10-port 1 port ethernet 1/7 link-group Itt ! card ge-10-port 1 port ethernet 1/8 link-group Itt ! card ge-10-port 1 port ethernet 1/9 link-group Itt ! card ge-10-port 1 port ethernet 1/10 link-group Itt ! -------------- next part -------------- An HTML attachment was scrubbed... URL: From shefys at gmail.com Fri May 13 18:35:21 2011 From: shefys at gmail.com (Yuri Shefer) Date: Fri, 13 May 2011 15:35:21 -0700 Subject: [rbak-nsp] Link-Group Questions In-Reply-To: References: Message-ID: Hi, Did you tried to use layer-4 balancing? [local]Redback(config)#service load-balance ip layer-4 On Fri, May 13, 2011 at 7:54 AM, Michael J. Gage wrote: > I have two Redback SE800 routers, and I am using a 4 port link-group between > them. > > > > Edge Router (hostname: Itt - SEOS-6.1.3.6-Release) using 2x XCRP3 & 4x > ge-4-port. > > Subscriber Router (hostname: Thing - - SEOS-6.1.5.7p5-Release) using 2x > XCRP4 & 1x ge-10-port. > > > > This issue is load-balancing across all 4 links. One direction I balance > across all 4 and the other it only uses 2. > > > > Edge ? Subscriber: > > ? 1. 1/3?????????????? (Up? )| 23.71% > > ??2. 2/3?????????????? (Up? )| 19.89% > > ??3. 3/3?????????????? (Up? )|? 0.00% > > ??4. 4/3?????????????? (Up? )|? 0.00% > > > > Subscriber ? Edge: > > ? 1. 1/7?????????????? (Up? )|? 2.60% > > ??2. 1/8?????????????? (Up? )|? 1.85% > > ??3. 1/9?????????????? (Up? )|? 2.75% > > ??4. 1/10????????????? (Up? )|? 2.25% > > > > Any thoughts? > -- Best regards, Yuri From mgage at localtel.net Sat May 14 10:17:20 2011 From: mgage at localtel.net (Michael J. Gage) Date: Sat, 14 May 2011 14:17:20 +0000 Subject: [rbak-nsp] Link-Group Questions In-Reply-To: References: Message-ID: I enabled the layer-4 option, and it had no effect on my port traffic. During peak data usage, I get dangerously close to maxing out the 2 ports. If I do not find an answer, I will have to split the 4 ports into 2 separate link-groups adding unwanted complexity to my network layout. -----Original Message----- From: Yuri Shefer [mailto:shefys at gmail.com] Sent: Friday, May 13, 2011 3:35 PM To: Michael J. Gage Cc: redback-nsp at puck.nether.net Subject: Re: [rbak-nsp] Link-Group Questions Hi, Did you tried to use layer-4 balancing? [local]Redback(config)#service load-balance ip layer-4 On Fri, May 13, 2011 at 7:54 AM, Michael J. Gage wrote: > I have two Redback SE800 routers, and I am using a 4 port link-group > between them. > > > > Edge Router (hostname: Itt - SEOS-6.1.3.6-Release) using 2x XCRP3 & 4x > ge-4-port. > > Subscriber Router (hostname: Thing - - SEOS-6.1.5.7p5-Release) using > 2x > XCRP4 & 1x ge-10-port. > > > > This issue is load-balancing across all 4 links. One direction I > balance across all 4 and the other it only uses 2. > > > > Edge ? Subscriber: > > ? 1. 1/3?????????????? (Up? )| 23.71% > > ??2. 2/3?????????????? (Up? )| 19.89% > > ??3. 3/3?????????????? (Up? )|? 0.00% > > ??4. 4/3?????????????? (Up? )|? 0.00% > > > > Subscriber ? Edge: > > ? 1. 1/7?????????????? (Up? )|? 2.60% > > ??2. 1/8?????????????? (Up? )|? 1.85% > > ??3. 1/9?????????????? (Up? )|? 2.75% > > ??4. 1/10????????????? (Up? )|? 2.25% > > > > Any thoughts? > -- Best regards, Yuri From layi.salako at gmail.com Mon May 16 06:12:36 2011 From: layi.salako at gmail.com (Oke-Salako Olayiwola) Date: Mon, 16 May 2011 11:12:36 +0100 Subject: [rbak-nsp] Ericsson Redback Message-ID: Hi Friends, Can someone help out with this? *Currently I use the Redback BRAS for postpaid ADSL service, I am using two RADIUS servers and in the Redback, I use tow contexts "local" and "ISP", context local is for the RADIUS servers and context ISP for the ADSL subscribers. I am about to start prepaid ADSL service and I have created a new context (context HSI) for the prepaid RADIUS servers, what I want to know now is do I create another context for the prepaid service and if I have to how do I make the users use the prepaid RADIUS servers.* *Note that I meant the initial configurations for postpaid service were done before I joined the coy * * * * * -- Thanks for reading my mail. Please reply. Oke-Salako Olayiwola +234-805-260-3665 +234-805-557-0604 -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.freedman at uk.clara.net Mon May 16 06:22:20 2011 From: david.freedman at uk.clara.net (David Freedman) Date: Mon, 16 May 2011 10:22:20 +0000 Subject: [rbak-nsp] Ericsson Redback In-Reply-To: Message-ID: The redback can use radius servers reachable from within the context or via another named context on the box (usually the local context), we call the second method "Global" AAA , you can choose which mechanism you use through use of the "aaa authentication" directive in the context, I.e: ! ! Use RADIUS server reachable from HSI ! context HSI aaa authentication subscriber radius radius server 1.2.3.4 key mykey ! OR ! ! Use RADIUS server reachable from Local Context ! context local radius server 1.2.3.4 key mykey ! context HSI aaa authentication subscriber global ! aaa global authentication subscriber radius context local ! Is there any particular reason you need to put your prepaid users in another context? There are sufficient mechanisms on this box to create a prepaid environment for users when they share a context with others (I.e radius driven ACLs, walled-gardens etc..), unless you have unique routing requirements or a completely different radius platform I would stay from using another context. Dave. From: Oke-Salako Olayiwola Date: Mon, 16 May 2011 11:12:36 +0100 To: Subject: [rbak-nsp] Ericsson Redback Hi Friends,Can someone help out with this? Currently I use the Redback BRAS for postpaid ADSL service, I am using two RADIUS servers and in the Redback, I use tow contexts "local" and "ISP", context local is for the RADIUS servers and context ISP for the ADSL subscribers. I am about to start prepaid ADSL service and I have created a new context (context HSI) for the prepaid RADIUS servers, what I want to know now is do I create another context for the prepaid service and if I have to how do I make the users use the prepaid RADIUS servers.Note that I meant the initial configurations for postpaid service were done before I joined the coy -- Thanks for reading my mail. Please reply. Oke-Salako Olayiwola +234-805-260-3665 +234-805-557-0604 _______________________________________________ redback-nsp mailing list redback-nsp at puck.nether.nethttps://puck.nether.net/mailman/listinfo/redback -nsp From layi.salako at gmail.com Mon May 16 06:37:47 2011 From: layi.salako at gmail.com (Oke-Salako Olayiwola) Date: Mon, 16 May 2011 11:37:47 +0100 Subject: [rbak-nsp] Ericsson Redback In-Reply-To: References: Message-ID: Thanks Dave, I am using different radius servers for the prepaid and postpaid subscribers, currently the postpaid subscribers connect using a domain created in context ISP "gloonline" (the username defined in the modem as xxxxx at gloonline). If I define the radius servers within the context (HSI), i believe if I create a domain the the same context, lets say domain "prepaidglo" and configure ADSL modems to use this xxxx at prepaidglo. it should work right? On 16 May 2011 11:22, David Freedman wrote: > The redback can use radius servers reachable from within the context or > via another named context on the box (usually the local context), we call > the second method "Global" AAA , you can choose which mechanism you use > through use of the "aaa authentication" directive in the context, I.e: > > ! > ! Use RADIUS server reachable from HSI > ! > context HSI > aaa authentication subscriber radius > radius server 1.2.3.4 key mykey > > ! > > > OR > > ! > ! Use RADIUS server reachable from Local Context > ! > context local > radius server 1.2.3.4 key mykey > > ! > context HSI > aaa authentication subscriber global > ! > aaa global authentication subscriber radius context local > ! > > > > > Is there any particular reason you need to put your prepaid users in > another context? There are sufficient mechanisms on this box to create a > prepaid environment for users when they share a context with others (I.e > radius driven ACLs, walled-gardens etc..), unless you have unique routing > requirements or a completely different radius platform I would stay from > using another context. > > Dave. > > > > From: Oke-Salako Olayiwola > Date: Mon, 16 May 2011 11:12:36 +0100 > To: > Subject: [rbak-nsp] Ericsson Redback > > > Hi Friends,Can someone help out with this? > > Currently I use the Redback BRAS for postpaid ADSL service, I am using two > RADIUS servers and in the Redback, I use tow contexts "local" and "ISP", > context local is for the RADIUS servers and context ISP for the ADSL > subscribers. I am about to start prepaid ADSL service and I have created a > new context (context HSI) for the prepaid RADIUS servers, what I want to > know now is do I create another context for the prepaid service and if I > have to how do I make the users use the prepaid RADIUS servers.Note that I > meant the initial configurations for postpaid service were done before I > joined the coy > > > > > > -- > Thanks for reading my mail. > Please reply. > > Oke-Salako Olayiwola > +234-805-260-3665 > +234-805-557-0604 > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.nethttps:// > puck.nether.net/mailman/listinfo/redback > -nsp > > -- Thanks for reading my mail. Please reply. Oke-Salako Olayiwola +234-805-260-3665 +234-805-557-0604 -------------- next part -------------- An HTML attachment was scrubbed... URL: From david.freedman at uk.clara.net Mon May 16 06:49:50 2011 From: david.freedman at uk.clara.net (David Freedman) Date: Mon, 16 May 2011 10:49:50 +0000 Subject: [rbak-nsp] Ericsson Redback In-Reply-To: References: , Message-ID: <66407E9F-8E60-426D-B848-55C520B103B0@uk.clara.net> On 16 May 2011, at 11:37, "Oke-Salako Olayiwola" > wrote: Thanks Dave, I am using different radius servers for the prepaid and postpaid subscribers, currently the postpaid subscribers connect using a domain created in context ISP "gloonline" (the username defined in the modem as xxxxx at gloonline). If I define the radius servers within the context (HSI), i believe if I create a domain the the same context, lets say domain "prepaidglo" and configure ADSL modems to use this xxxx at prepaidglo. it should work right? What kind of subscriber's are these? Direct PPP? L2TP? Depends on how subscribers are getting mapped to contexts and if mapping via this new domain and context is allowed for them, but yes, in theory if they can select this context via the domain and this context has it's own, reachable radius servers then this should be fine On 16 May 2011 11:22, David Freedman <david.freedman at uk.clara.net> wrote: The redback can use radius servers reachable from within the context or via another named context on the box (usually the local context), we call the second method "Global" AAA , you can choose which mechanism you use through use of the "aaa authentication" directive in the context, I.e: ! ! Use RADIUS server reachable from HSI ! context HSI aaa authentication subscriber radius radius server 1.2.3.4 key mykey ! OR ! ! Use RADIUS server reachable from Local Context ! context local radius server 1.2.3.4 key mykey ! context HSI aaa authentication subscriber global ! aaa global authentication subscriber radius context local ! Is there any particular reason you need to put your prepaid users in another context? There are sufficient mechanisms on this box to create a prepaid environment for users when they share a context with others (I.e radius driven ACLs, walled-gardens etc..), unless you have unique routing requirements or a completely different radius platform I would stay from using another context. Dave. From: Oke-Salako Olayiwola <layi.salako at gmail.com> Date: Mon, 16 May 2011 11:12:36 +0100 To: <redback-nsp at puck.nether.net> Subject: [rbak-nsp] Ericsson Redback Hi Friends,Can someone help out with this? Currently I use the Redback BRAS for postpaid ADSL service, I am using two RADIUS servers and in the Redback, I use tow contexts "local" and "ISP", context local is for the RADIUS servers and context ISP for the ADSL subscribers. I am about to start prepaid ADSL service and I have created a new context (context HSI) for the prepaid RADIUS servers, what I want to know now is do I create another context for the prepaid service and if I have to how do I make the users use the prepaid RADIUS servers.Note that I meant the initial configurations for postpaid service were done before I joined the coy -- Thanks for reading my mail. Please reply. Oke-Salako Olayiwola +234-805-260-3665 +234-805-557-0604 _______________________________________________ redback-nsp mailing list redback-nsp at puck.nether.nethttps://puck.nether.net/mailman/listinfo/redback -nsp -- Thanks for reading my mail. Please reply. Oke-Salako Olayiwola +234-805-260-3665 +234-805-557-0604 -------------- next part -------------- An HTML attachment was scrubbed... URL: From layi.salako at gmail.com Tue May 17 10:45:14 2011 From: layi.salako at gmail.com (Oke-Salako Olayiwola) Date: Tue, 17 May 2011 15:45:14 +0100 Subject: [rbak-nsp] Ericsson Redback In-Reply-To: <66407E9F-8E60-426D-B848-55C520B103B0@uk.clara.net> References: <66407E9F-8E60-426D-B848-55C520B103B0@uk.clara.net> Message-ID: On 16 May 2011 11:49, David Freedman wrote: > > > On 16 May 2011, at 11:37, "Oke-Salako Olayiwola" > wrote: > > Thanks Dave, I am using different radius servers for the prepaid and > postpaid subscribers, currently the postpaid subscribers connect using a > domain created in context ISP "gloonline" (the username defined in the modem > as xxxxx at gloonline). If I define the radius servers within the context > (HSI), i believe if I create a domain the the same context, lets say domain > "prepaidglo" and configure ADSL modems to use this xxxx at prepaidglo. it > should work right? > > the subscribers are PPP > > What kind of subscriber's are these? Direct PPP? L2TP? Depends on how > subscribers are getting mapped to contexts and if mapping via this new > domain and context is allowed for them, but yes, in theory if they can > select this context via the domain and this context has it's own, reachable > radius servers then this should be fine > > > On 16 May 2011 11:22, David Freedman < > david.freedman at uk.clara.net> wrote: > >> The redback can use radius servers reachable from within the context or >> via another named context on the box (usually the local context), we call >> the second method "Global" AAA , you can choose which mechanism you use >> through use of the "aaa authentication" directive in the context, I.e: >> >> ! >> ! Use RADIUS server reachable from HSI >> ! >> context HSI >> aaa authentication subscriber radius >> radius server 1.2.3.4 key mykey >> >> ! >> >> >> OR >> >> ! >> ! Use RADIUS server reachable from Local Context >> ! >> context local >> radius server 1.2.3.4 key mykey >> >> ! >> context HSI >> aaa authentication subscriber global >> ! >> aaa global authentication subscriber radius context local >> ! >> >> >> >> >> Is there any particular reason you need to put your prepaid users in >> another context? There are sufficient mechanisms on this box to create a >> prepaid environment for users when they share a context with others (I.e >> radius driven ACLs, walled-gardens etc..), unless you have unique routing >> requirements or a completely different radius platform I would stay from >> using another context. >> >> Dave. >> >> >> >> From: Oke-Salako Olayiwola < >> layi.salako at gmail.com> >> Date: Mon, 16 May 2011 11:12:36 +0100 >> To: < redback-nsp at puck.nether.net> >> Subject: [rbak-nsp] Ericsson Redback >> >> >> Hi Friends,Can someone help out with this? >> >> Currently I use the Redback BRAS for postpaid ADSL service, I am using two >> RADIUS servers and in the Redback, I use tow contexts "local" and "ISP", >> context local is for the RADIUS servers and context ISP for the ADSL >> subscribers. I am about to start prepaid ADSL service and I have created a >> new context (context HSI) for the prepaid RADIUS servers, what I want to >> know now is do I create another context for the prepaid service and if I >> have to how do I make the users use the prepaid RADIUS servers.Note that I >> meant the initial configurations for postpaid service were done before I >> joined the coy >> >> >> >> >> >> -- >> Thanks for reading my mail. >> Please reply. >> >> Oke-Salako Olayiwola >> +234-805-260-3665 >> +234-805-557-0604 >> >> _______________________________________________ >> redback-nsp mailing list >> redback-nsp at puck.nether.nethttps:// >> puck.nether.net/mailman/listinfo/redback >> -nsp >> >> > > > -- > Thanks for reading my mail. > Please reply. > > Oke-Salako Olayiwola > +234-805-260-3665 > +234-805-557-0604 > > -- Thanks for reading my mail. Please reply. Oke-Salako Olayiwola +234-805-260-3665 +234-805-557-0604 -------------- next part -------------- An HTML attachment was scrubbed... URL: From pj at hostersi.pl Tue May 17 15:37:59 2011 From: pj at hostersi.pl (Pawel Jarosz) Date: Tue, 17 May 2011 21:37:59 +0200 (CEST) Subject: [rbak-nsp] Source based routing on SE100 Message-ID: Hi, is there a way to route traffic from specific IP addresses to context A and other IP addresses to context B? Is this possible? Any hints are welcome. Regards, Pawel From frans at falco-networks.com Wed May 18 08:05:14 2011 From: frans at falco-networks.com (Frans Legdeur) Date: Wed, 18 May 2011 14:05:14 +0200 Subject: [rbak-nsp] Source based routing on SE100 In-Reply-To: Message-ID: What about per mac adres in a different context? Would that work Pawel? This would be possible using (static/dynamic) clips, as long as these are subscribers. Else you would need intercontext routing with route filters and ACL's. Kind regards, Frans. > From: Pawel Jarosz > Date: Tue, 17 May 2011 21:37:59 +0200 (CEST) > To: > Subject: [rbak-nsp] Source based routing on SE100 > > Hi, > is there a way to route > traffic from specific IP addresses to context A > and other IP addresses to context B? > Is this possible? > > Any hints are welcome. > > Regards, > Pawel > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp From pj at hostersi.pl Wed May 18 08:46:53 2011 From: pj at hostersi.pl (Pawel Jarosz) Date: Wed, 18 May 2011 14:46:53 +0200 (CEST) Subject: [rbak-nsp] Source based routing on SE100 In-Reply-To: References: Message-ID: On Wed, 18 May 2011, Frans Legdeur wrote: > What about per mac adres in a different context? > Would that work Pawel? > > This would be possible using (static/dynamic) clips, as long as these are > subscribers. This border router, so there is no l2 information. With static clips it partially works, works towards context A or B but it does not work backwards (bellow sample config): context ospf interface klips multibind ip unnumbered ospf-area ip pool 10.0.0.0/16 interface ospf-area description OSPF core member ip address 10.11.0.0/24 subscriber name s2 ip address 10.0.0.5 forward policy fwd-b ip route 0.0.0.0/0 context A ! forward policy fwd-b redirect destination next-hop ip-of-context-b ! port ethernet 2/1 bind interface ospf-area ospf service clips clips pvc 1 bind subscriber s1 at ospf Now I have working ospf, working forward subscriber, but in routing table there is entry: Type Network Next Hop Dist Metric UpTime Interface SUB A 10.0.0.5/32 10.0.0.5 15 0 klips not what I have expected. Is there a way to supress this route? I do not mean to configure it by hand in config - this is not flexible. > Else you would need intercontext routing with route filters and ACL's. Could you provide an example? Than you, Pawel From prapti.soe at gmail.com Wed May 18 11:20:03 2011 From: prapti.soe at gmail.com (soe prapti) Date: Wed, 18 May 2011 22:20:03 +0700 Subject: [rbak-nsp] user drop slowly when using authentication none Message-ID: Dear All I use se1200 with 3 card 4 x 10G and 1 card 5 x 1G. The AAA using external radius . Because of the external radius sometimes dead then queing in accounting server is very high ( status of radius accounting is FULL) so we are using "aaa authentication subscriber none" to keep the number of subscriber. Ussually subscriber active is 55000 user. When we are using authentication subscriber none and number subscriber active 55000 immediatelly the subscriber connection was drop slowy, and in the show log we find so many information like this example : May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/25731 May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131183116514 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: AAA session alloc failed for cct 3/3:1023:63/6/2/6112 May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131173104855 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/195742 May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131183100681 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/102520 May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131184105822 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/71590 May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131184109151 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/47718 May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131183115892 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/118063 May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131182105848 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: AAA session alloc failed for cct 3/3:1023:63/6/2/229398 May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131161102087 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/89699 May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131183108829 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: AAA session alloc failed for cct 3/3:1023:63/6/2/30327 May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131159101083 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/184331 May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131167109587 at telkom.net May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: AAA session alloc failed for cct 1/3:1023:63/6/2/36613 May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: Failed to allocate aaa_session structure for username 131184108694 at telkom.net Any body can help me? Thanks -- Regards, Suparti 081384850602 -------------- next part -------------- An HTML attachment was scrubbed... URL: From ian at calderbankconsulting.co.uk Wed May 18 12:29:56 2011 From: ian at calderbankconsulting.co.uk (Ian Calderbank) Date: Wed, 18 May 2011 17:29:56 +0100 Subject: [rbak-nsp] Source based routing on SE100 Message-ID: <002401cc1578$d8ff98b0$8afeca10$@calderbankconsulting.co.uk> Pawel, Its not clear to me what you are asking / trying to do, the explanation seems a little confused? If you are asking, whats the equivalent of cisco PBR (make forwarding decision based on source IP or other arbitrary criteria), its "forwarding policy" applied to circuit, which it seems you are already trying using. You can apply forwarding policy by radius attribute per-subscriber, if that's more scalable for you. Most scalable design is to bind the subscriber to the right context in the first place. Why are you binding all subs to context OSPF? Why not bind him to context A in the first place, then he automatically routes through context A? The subscriber static route that you see is expected behaviour given that you have defined that static subscriber and it is bound. cheers Ian ---------- 1. Source based routing on SE100 (Pawel Jarosz) 2. Re: Source based routing on SE100 (Frans Legdeur) 3. Re: Source based routing on SE100 (Pawel Jarosz) Message: 3 Date: Wed, 18 May 2011 14:46:53 +0200 (CEST) From: Pawel Jarosz To: Frans Legdeur Cc: redback-nsp at puck.nether.net Subject: Re: [rbak-nsp] Source based routing on SE100 Message-ID: Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII On Wed, 18 May 2011, Frans Legdeur wrote: > What about per mac adres in a different context? > Would that work Pawel? > > This would be possible using (static/dynamic) clips, as long as these > are subscribers. This border router, so there is no l2 information. With static clips it partially works, works towards context A or B but it does not work backwards (bellow sample config): context ospf interface klips multibind ip unnumbered ospf-area ip pool 10.0.0.0/16 interface ospf-area description OSPF core member ip address 10.11.0.0/24 subscriber name s2 ip address 10.0.0.5 forward policy fwd-b ip route 0.0.0.0/0 context A ! forward policy fwd-b redirect destination next-hop ip-of-context-b ! port ethernet 2/1 bind interface ospf-area ospf service clips clips pvc 1 bind subscriber s1 at ospf Now I have working ospf, working forward subscriber, but in routing table there is entry: Type Network Next Hop Dist Metric UpTime Interface SUB A 10.0.0.5/32 10.0.0.5 15 0 klips not what I have expected. Is there a way to supress this route? I do not mean to configure it by hand in config - this is not flexible. > Else you would need intercontext routing with route filters and ACL's. Could you provide an example? Than you, Pawel ------------------------------ From pj at hostersi.pl Wed May 18 13:22:06 2011 From: pj at hostersi.pl (Pawel Jarosz) Date: Wed, 18 May 2011 19:22:06 +0200 (CEST) Subject: [rbak-nsp] Source based routing on SE100 In-Reply-To: <002401cc1578$d8ff98b0$8afeca10$@calderbankconsulting.co.uk> References: <002401cc1578$d8ff98b0$8afeca10$@calderbankconsulting.co.uk> Message-ID: On Wed, 18 May 2011, Ian Calderbank wrote: > Pawel, Ian, > Its not clear to me what you are asking / trying to do, the explanation > seems a little confused? > > If you are asking, whats the equivalent of cisco PBR (make forwarding > decision based on source IP or other arbitrary criteria), its "forwarding > policy" applied to circuit, which it seems you are already trying using. Exactly. > You can apply forwarding policy by radius attribute per-subscriber, if > that's more scalable for you. Static is for now, I try to get it work. > Most scalable design is to bind the subscriber to the right context in the > first place. Why are you binding all subs to context OSPF? Why not bind him > to context A in the first place, then he automatically routes through > context A? topology is: customer(subscriber)- (port) --OSPF --- (port) - ISP A (bgp) |-- (port) - ISP B (BGP) > The subscriber static route that you see is expected behaviour given that > you have defined that static subscriber and it is bound. Maybe I get it wrong, and clips is not for routed subscribers, only direct attached? If so, the route is created as expected. Pawel From gigaeth0 at gmail.com Wed May 18 21:36:39 2011 From: gigaeth0 at gmail.com (Giga Eth0) Date: Thu, 19 May 2011 11:36:39 +1000 Subject: [rbak-nsp] Ericsson Redback In-Reply-To: References: <66407E9F-8E60-426D-B848-55C520B103B0@uk.clara.net> Message-ID: Just out of curiosity, why are you using separate Radius servers for your prepaid and postpaid users? Wouldn't it be easier, efficient and cheaper to have all subscribers on the same Radius server? As Dave has pointed out, you can still have both your Radius servers in the local context and have the subscriber terminate into separate contexts for prepaid and postpaid. //jhq On Wed, May 18, 2011 at 12:45 AM, Oke-Salako Olayiwola < layi.salako at gmail.com> wrote: > > > On 16 May 2011 11:49, David Freedman wrote: > >> >> >> On 16 May 2011, at 11:37, "Oke-Salako Olayiwola" >> wrote: >> >> Thanks Dave, I am using different radius servers for the prepaid and >> postpaid subscribers, currently the postpaid subscribers connect using a >> domain created in context ISP "gloonline" (the username defined in the modem >> as xxxxx at gloonline). If I define the radius servers within the context >> (HSI), i believe if I create a domain the the same context, lets say domain >> "prepaidglo" and configure ADSL modems to use this xxxx at prepaidglo. it >> should work right? >> >> the subscribers are PPP >> > > >> What kind of subscriber's are these? Direct PPP? L2TP? Depends on how >> subscribers are getting mapped to contexts and if mapping via this new >> domain and context is allowed for them, but yes, in theory if they can >> select this context via the domain and this context has it's own, reachable >> radius servers then this should be fine >> >> >> On 16 May 2011 11:22, David Freedman < >> david.freedman at uk.clara.net> wrote: >> >>> The redback can use radius servers reachable from within the context or >>> via another named context on the box (usually the local context), we call >>> the second method "Global" AAA , you can choose which mechanism you use >>> through use of the "aaa authentication" directive in the context, I.e: >>> >>> ! >>> ! Use RADIUS server reachable from HSI >>> ! >>> context HSI >>> aaa authentication subscriber radius >>> radius server 1.2.3.4 key mykey >>> >>> ! >>> >>> >>> OR >>> >>> ! >>> ! Use RADIUS server reachable from Local Context >>> ! >>> context local >>> radius server 1.2.3.4 key mykey >>> >>> ! >>> context HSI >>> aaa authentication subscriber global >>> ! >>> aaa global authentication subscriber radius context local >>> ! >>> >>> >>> >>> >>> Is there any particular reason you need to put your prepaid users in >>> another context? There are sufficient mechanisms on this box to create a >>> prepaid environment for users when they share a context with others (I.e >>> radius driven ACLs, walled-gardens etc..), unless you have unique routing >>> requirements or a completely different radius platform I would stay from >>> using another context. >>> >>> Dave. >>> >>> >>> >>> From: Oke-Salako Olayiwola < >>> layi.salako at gmail.com> >>> Date: Mon, 16 May 2011 11:12:36 +0100 >>> To: < redback-nsp at puck.nether.net> >>> Subject: [rbak-nsp] Ericsson Redback >>> >>> >>> Hi Friends,Can someone help out with this? >>> >>> Currently I use the Redback BRAS for postpaid ADSL service, I am using >>> two >>> RADIUS servers and in the Redback, I use tow contexts "local" and "ISP", >>> context local is for the RADIUS servers and context ISP for the ADSL >>> subscribers. I am about to start prepaid ADSL service and I have created >>> a >>> new context (context HSI) for the prepaid RADIUS servers, what I want to >>> know now is do I create another context for the prepaid service and if I >>> have to how do I make the users use the prepaid RADIUS servers.Note that >>> I >>> meant the initial configurations for postpaid service were done before I >>> joined the coy >>> >>> >>> >>> >>> >>> -- >>> Thanks for reading my mail. >>> Please reply. >>> >>> Oke-Salako Olayiwola >>> +234-805-260-3665 >>> +234-805-557-0604 >>> >>> _______________________________________________ >>> redback-nsp mailing list >>> redback-nsp at puck.nether.nethttps:// >>> puck.nether.net/mailman/listinfo/redback >>> -nsp >>> >>> >> >> >> -- >> Thanks for reading my mail. >> Please reply. >> >> Oke-Salako Olayiwola >> +234-805-260-3665 >> +234-805-557-0604 >> >> > > > -- > Thanks for reading my mail. > Please reply. > > Oke-Salako Olayiwola > +234-805-260-3665 > +234-805-557-0604 > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gigaeth0 at gmail.com Wed May 18 23:14:07 2011 From: gigaeth0 at gmail.com (Giga Eth0) Date: Thu, 19 May 2011 13:14:07 +1000 Subject: [rbak-nsp] user drop slowly when using authentication none In-Reply-To: References: Message-ID: What subscriber license are you using on your SmartEdge? Can you paste a "show licenses" please. On Thu, May 19, 2011 at 1:20 AM, soe prapti wrote: > Dear All > > I use se1200 with 3 card 4 x 10G and 1 card 5 x 1G. The AAA using external > radius . Because of the external radius sometimes dead then queing in > accounting server is very high ( status of radius accounting is FULL) so we > are using "aaa authentication subscriber none" to keep the number of > subscriber. Ussually subscriber active is 55000 user. When we are using > authentication subscriber none and number subscriber active 55000 > immediatelly the subscriber connection was drop slowy, and in the show log > we find so many information like this example : > > > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/25731 > May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131183116514 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: AAA session alloc > failed for cct 3/3:1023:63/6/2/6112 > May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131173104855 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/195742 > May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131183100681 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/102520 > May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131184105822 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/71590 > May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131184109151 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/47718 > May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131183115892 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/118063 > May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131182105848 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: AAA session alloc > failed for cct 3/3:1023:63/6/2/229398 > May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131161102087 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/89699 > May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131183108829 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: AAA session alloc > failed for cct 3/3:1023:63/6/2/30327 > May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131159101083 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/184331 > May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131167109587 at telkom.net > May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed > May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: AAA session alloc > failed for cct 1/3:1023:63/6/2/36613 > May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: Failed to allocate > aaa_session structure for username 131184108694 at telkom.net > > > Any body can help me? Thanks > -- > Regards, > > Suparti > 081384850602 > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From prapti.soe at gmail.com Wed May 18 23:18:16 2011 From: prapti.soe at gmail.com (soe prapti) Date: Thu, 19 May 2011 10:18:16 +0700 Subject: [rbak-nsp] user drop slowly when using authentication none In-Reply-To: References: Message-ID: Hi Giga, The license is like this [local]BRAS-LBG-SNT#show license Software Feature License Configured -------------------------- ------------------ subscriber dynamic-service YES l2tp all YES mpls YES subscriber high-availability YES subscriber active 96000 YES subscriber bandwidth 1000 YES Total active subscriber license configured 96000 thanks On Thu, May 19, 2011 at 10:14 AM, Giga Eth0 wrote: > What subscriber license are you using on your SmartEdge? Can you paste a > "show licenses" please. > > On Thu, May 19, 2011 at 1:20 AM, soe prapti wrote: > >> Dear All >> >> I use se1200 with 3 card 4 x 10G and 1 card 5 x 1G. The AAA using external >> radius . Because of the external radius sometimes dead then queing in >> accounting server is very high ( status of radius accounting is FULL) so we >> are using "aaa authentication subscriber none" to keep the number of >> subscriber. Ussually subscriber active is 55000 user. When we are using >> authentication subscriber none and number subscriber active 55000 >> immediatelly the subscriber connection was drop slowy, and in the show log >> we find so many information like this example : >> >> >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/25731 >> May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131183116514 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: AAA session alloc >> failed for cct 3/3:1023:63/6/2/6112 >> May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131173104855 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/195742 >> May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131183100681 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/102520 >> May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131184105822 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/71590 >> May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131184109151 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/47718 >> May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131183115892 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/118063 >> May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131182105848 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: AAA session alloc >> failed for cct 3/3:1023:63/6/2/229398 >> May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131161102087 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/89699 >> May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131183108829 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: AAA session alloc >> failed for cct 3/3:1023:63/6/2/30327 >> May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131159101083 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/184331 >> May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131167109587 at telkom.net >> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >> May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: AAA session alloc >> failed for cct 1/3:1023:63/6/2/36613 >> May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: Failed to allocate >> aaa_session structure for username 131184108694 at telkom.net >> >> >> Any body can help me? Thanks >> -- >> Regards, >> >> Suparti >> 081384850602 >> >> _______________________________________________ >> redback-nsp mailing list >> redback-nsp at puck.nether.net >> https://puck.nether.net/mailman/listinfo/redback-nsp >> >> > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > > -- Regards, Suparti 081384850602 -------------- next part -------------- An HTML attachment was scrubbed... URL: From gigaeth0 at gmail.com Wed May 18 23:35:45 2011 From: gigaeth0 at gmail.com (Giga Eth0) Date: Thu, 19 May 2011 13:35:45 +1000 Subject: [rbak-nsp] Source based routing on SE100 In-Reply-To: References: <002401cc1578$d8ff98b0$8afeca10$@calderbankconsulting.co.uk> Message-ID: I agree that a much cleaner solution is to map a subscriber to a context directly. But for other cases (assuming non subscriber traffic), PBR is possible. Here is a snippet for source based routing.. policy access-list Source_based_routing seq 10 permit ip 192.168.1.0 0.0.255.255 any class ClassA seq 20 permit ip 192.168.2.0 0.0.255.255 any class ClassB seq 30 permit ip 192.168.3.0 0.0.255.255 any class ClassC forward policy Source_based_redirect access-group Source_based_routing class ClassA redirect destination next-hop 10.0.0.1 class ClassB redirect destination next-hop 10.0.1.1 class ClassC redirect destination next-hop 10.0.2.1 //cheers On Thu, May 19, 2011 at 3:22 AM, Pawel Jarosz wrote: > On Wed, 18 May 2011, Ian Calderbank wrote: > > Pawel, >> > Ian, > > > Its not clear to me what you are asking / trying to do, the explanation >> seems a little confused? >> >> If you are asking, whats the equivalent of cisco PBR (make forwarding >> decision based on source IP or other arbitrary criteria), its "forwarding >> policy" applied to circuit, which it seems you are already trying using. >> > Exactly. > > > You can apply forwarding policy by radius attribute per-subscriber, if >> that's more scalable for you. >> > Static is for now, I try to get it work. > > > Most scalable design is to bind the subscriber to the right context in the >> first place. Why are you binding all subs to context OSPF? Why not bind >> him >> to context A in the first place, then he automatically routes through >> context A? >> > topology is: > customer(subscriber)- (port) --OSPF --- (port) - ISP A (bgp) > |-- (port) - ISP B (BGP) > > > The subscriber static route that you see is expected behaviour given that >> you have defined that static subscriber and it is bound. >> > Maybe I get it wrong, and clips is not for > routed subscribers, only direct attached? > If so, the route is created as expected. > > > Pawel > > _______________________________________________ > redback-nsp mailing list > redback-nsp at puck.nether.net > https://puck.nether.net/mailman/listinfo/redback-nsp > -------------- next part -------------- An HTML attachment was scrubbed... URL: From layi.salako at gmail.com Thu May 19 06:05:16 2011 From: layi.salako at gmail.com (Olayiwola Oke-Salako) Date: Thu, 19 May 2011 10:05:16 +0000 (UTC) Subject: [rbak-nsp] Invitation to connect on LinkedIn Message-ID: <2032760792.21749310.1305799516881.JavaMail.app@ela4-bed35.prod> LinkedIn ------------ I'd like to add you to my professional network on LinkedIn. - Olayiwola Olayiwola Oke-Salako Training at Alcatel University Nigeria Confirm that you know Olayiwola Oke-Salako https://www.linkedin.com/e/lme90e-gnvjbfm6-3k/isd/2932524948/rM5gZB7M/ -- (c) 2011, LinkedIn Corporation -------------- next part -------------- An HTML attachment was scrubbed... URL: From member at linkedin.com Fri May 20 13:31:48 2011 From: member at linkedin.com (Marvin Panergo via LinkedIn) Date: Fri, 20 May 2011 17:31:48 +0000 (UTC) Subject: [rbak-nsp] Invitation to connect on LinkedIn Message-ID: <1199529819.19128597.1305912708299.JavaMail.app@ela4-bed84.prod> LinkedIn ------------Marvin Panergo requested to add you as a connection on LinkedIn: ------------------------------------------ Ronald, I'd like to add you to my professional network on LinkedIn. - Marvin Accept invitation from Marvin Panergo http://www.linkedin.com/e/lme90e-gnxepio9-6w/qXn2sTNbBiceXByA_FttVDNF1YsWey03kontXEQy_r/blk/I114785450_25/1BpC5vrmRLoRZcjkkZt5YCpnlOt3RApnhMpmdzgmhxrSNBszYRcBYMdjgRe3sQcj59bTEQhTl2k4RPbPsVejAUdjwQej8LrCBxbOYWrSlI/EML_comm_afe/ View invitation from Marvin Panergo http://www.linkedin.com/e/lme90e-gnxepio9-6w/qXn2sTNbBiceXByA_FttVDNF1YsWey03kontXEQy_r/blk/I114785450_25/3kOnP0Rd3kUdPgNckALqnpPbOYWrSlI/svi/ ------------------------------------------ DID YOU KNOW LinkedIn can help you find the right service providers using recommendations from your trusted network? Using LinkedIn Services, you can take the risky guesswork out of selecting service providers by reading the recommendations of credible, trustworthy members of your network. http://www.linkedin.com/e/lme90e-gnxepio9-6w/svp/inv-25/ -- (c) 2011, LinkedIn Corporation -------------- next part -------------- An HTML attachment was scrubbed... URL: From member at linkedin.com Fri May 20 13:35:06 2011 From: member at linkedin.com (Marvin Panergo via LinkedIn) Date: Fri, 20 May 2011 17:35:06 +0000 (UTC) Subject: [rbak-nsp] Invitation to connect on LinkedIn Message-ID: <200973068.25723351.1305912906885.JavaMail.app@ela4-bed78.prod> LinkedIn ------------Marvin Panergo requested to add you as a connection on LinkedIn: ------------------------------------------ Ronald, I'd like to add you to my professional network on LinkedIn. - Marvin Accept invitation from Marvin Panergo http://www.linkedin.com/e/lme90e-gnxetrwi-4u/qXn2sTNbBiceXByA_FttVDNF1YsWey03kontXEQy_r/blk/I114787209_25/1BpC5vrmRLoRZcjkkZt5YCpnlOt3RApnhMpmdzgmhxrSNBszYRcBYVc38Te3sQcj59bTEQhTl2k4RPbPsVejAUdjwQej8LrCBxbOYWrSlI/EML_comm_afe/ View invitation from Marvin Panergo http://www.linkedin.com/e/lme90e-gnxetrwi-4u/qXn2sTNbBiceXByA_FttVDNF1YsWey03kontXEQy_r/blk/I114787209_25/3kOnPAMczsUdPgNckALqnpPbOYWrSlI/svi/ ------------------------------------------ DID YOU KNOW LinkedIn can help you find the right service providers using recommendations from your trusted network? Using LinkedIn Services, you can take the risky guesswork out of selecting service providers by reading the recommendations of credible, trustworthy members of your network. http://www.linkedin.com/e/lme90e-gnxetrwi-4u/svp/inv-25/ -- (c) 2011, LinkedIn Corporation -------------- next part -------------- An HTML attachment was scrubbed... URL: From gigaeth0 at gmail.com Mon May 23 02:07:16 2011 From: gigaeth0 at gmail.com (Giga Eth0) Date: Mon, 23 May 2011 16:07:16 +1000 Subject: [rbak-nsp] user drop slowly when using authentication none In-Reply-To: References: Message-ID: This is unexpected behavior. Can you check on your AAA if the subscriber is being authenticated even before the 55000 limit is reached. It might be a case of mis-configuration and the users still being authenticated via the AAA. Could you also paste your AAA config from the Redback here. cheers On Thu, May 19, 2011 at 1:18 PM, soe prapti wrote: > Hi Giga, > > > The license is like this > > > [local]BRAS-LBG-SNT#show license > Software Feature License Configured > -------------------------- ------------------ > subscriber dynamic-service YES > l2tp all YES > mpls YES > subscriber high-availability YES > subscriber active 96000 YES > subscriber bandwidth 1000 YES > > Total active subscriber license configured 96000 > > > thanks > > > > On Thu, May 19, 2011 at 10:14 AM, Giga Eth0 wrote: > >> What subscriber license are you using on your SmartEdge? Can you paste a >> "show licenses" please. >> >> On Thu, May 19, 2011 at 1:20 AM, soe prapti wrote: >> >>> Dear All >>> >>> I use se1200 with 3 card 4 x 10G and 1 card 5 x 1G. The AAA using >>> external radius . Because of the external radius sometimes dead then queing >>> in accounting server is very high ( status of radius accounting is FULL) so >>> we are using "aaa authentication subscriber none" to keep the number of >>> subscriber. Ussually subscriber active is 55000 user. When we are using >>> authentication subscriber none and number subscriber active 55000 >>> immediatelly the subscriber connection was drop slowy, and in the show log >>> we find so many information like this example : >>> >>> >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/25731 >>> May 17 17:14:12: [1/3:1023:63/6/2/25731]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131183116514 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: AAA session alloc >>> failed for cct 3/3:1023:63/6/2/6112 >>> May 17 17:14:12: [3/3:1023:63/6/2/6112]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131173104855 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/195742 >>> May 17 17:14:12: [1/3:1023:63/6/2/195742]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131183100681 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/102520 >>> May 17 17:14:12: [1/3:1023:63/6/2/102520]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131184105822 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/71590 >>> May 17 17:14:12: [1/3:1023:63/6/2/71590]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131184109151 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/47718 >>> May 17 17:14:12: [1/3:1023:63/6/2/47718]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131183115892 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/118063 >>> May 17 17:14:12: [1/3:1023:63/6/2/118063]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131182105848 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: AAA session alloc >>> failed for cct 3/3:1023:63/6/2/229398 >>> May 17 17:14:12: [3/3:1023:63/6/2/229398]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131161102087 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/89699 >>> May 17 17:14:12: [1/3:1023:63/6/2/89699]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131183108829 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: AAA session alloc >>> failed for cct 3/3:1023:63/6/2/30327 >>> May 17 17:14:12: [3/3:1023:63/6/2/30327]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131159101083 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/184331 >>> May 17 17:14:12: [1/3:1023:63/6/2/184331]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131167109587 at telkom.net >>> May 17 17:14:12: %AAA-3-ERR: AAA sm sesson alloc failed >>> May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: AAA session alloc >>> failed for cct 1/3:1023:63/6/2/36613 >>> May 17 17:14:12: [1/3:1023:63/6/2/36613]: %AAA-3-ERR: Failed to allocate >>> aaa_session structure for username 131184108694 at telkom.net >>> >>> >>> Any body can help me? Thanks >>> -- >>> Regards, >>> >>> Suparti >>> 081384850602 >>> >>> _______________________________________________ >>> redback-nsp mailing list >>> redback-nsp at puck.nether.net >>> https://puck.nether.net/mailman/listinfo/redback-nsp >>> >>> >> >> _______________________________________________ >> redback-nsp mailing list >> redback-nsp at puck.nether.net >> https://puck.nether.net/mailman/listinfo/redback-nsp >> >> > > > -- > Regards, > > Suparti > 081384850602 > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mloveley at gmail.com Thu May 26 05:31:59 2011 From: mloveley at gmail.com (Mark Loveley) Date: Thu, 26 May 2011 10:31:59 +0100 Subject: [rbak-nsp] 6.3.1.3 - Sending ipv4 vpn AF routes via BGP issue Message-ID: Hi I'm testing SEOS 6.3.1.3 in the lab and am seeing some issues with the advertisement of ipv4 vpn routes from the Smartedge. I wondered if anyone else has seen similar issues. After upgrade The session comes up. The unicast routes are sent, but the vpn AF routes are not. [local]Eeyore#sh bgp neighbor 192.168.102.6 BGP neighbor: 192.168.102.6, remote AS: 65000, internal link Version: 4, router identifier: 192.168.102.6 Peer Group member: internal State: Established for 22:33:30 Description: rabbitFakeEdge Last read 00:00:20, last send 00:00:22 Hold time: configured 180, negotiated 90 Keepalive time: configured 60, negotiated 30 Local restart timer 120 sec, stale route retain timer 180 sec Received restart timer 120 sec, flag 0x0 Minimum time between advertisement runs: 5 secs Source IP address used from interface: loopback0 Source (local) IP address: 192.168.102.1 Received messages: 2992 (56977 bytes), notifications: 0, in queue: 0 Sent messages: 2687 (51509 bytes), notifications: 0, out queue: 0 Reset count: 2, last reset time: 22:33:39, reset reason: User action CapSent: refresh, 4byteAS, unicast, vpn, restart CapRcvd: refresh, 4byteAS, unicast, vpn restart (time 120, flags 0x0, empty) Address family: ipv4 unicast Peer Group member: internal BGP table version: 47, neighbor version: 47 Routes: rcvd 0, imported 0, active 0, history 0, dampend 0, sent 12 Address family: ipv4 vpn Peer Group member: internal BGP table version: 184, neighbor version: 0 Routes: rcvd 1, imported 2, active 3, history 0, dampend 0, sent 0 [local]Eeyore# The Juniper on the other end of the BGP session never see's the "end-of-rib marker" for inet-vpn-unicast. Using BGP traceoptions (flag all) on the Juniper I can see that it never seems to receive the "BGP RECV End of RIB: AFI 1 SAFI 128", whereas it does receive the one for "SAFI 1", noc at Rabbit> show log fakeEdge/fakeEdge-BGP | match AFI May 18 12:57:10.579096 BGP SEND End of RIB: AFI 1 SAFI 1 May 18 12:57:10.579642 BGP SEND End of RIB: AFI 1 SAFI 128 May 18 12:57:10.584650 BGP RECV End of RIB: AFI 1 SAFI 1 {master} After some time ( ie 12+ hours) "an event" occurs (which I haven't tracked down yet, it's not regular) and the vpn AF routes start being sent. Once the routes are being sent, If I clear a BGP session after that event, it advertises the vpn routes straight away until the Smartedge is rebooted. After the reboot the vpn routes aren't sent. If I downgrade to 6.1.3.8p1 and the same config works first time every time. 13 routes are advertised in the vpn address family. This happens with all iBGP peers on the Smartedge, not just this one. There ARE routes locally within that AF ready to be sent, it's just not sending them [local]Eeyore#sh bgp route ipv4 vpn Address Family: ipv4 vpn BGP table version is 184, local router ID is 192.168.102.1 Status codes: d damped, h history, > best, i internal Origin codes: i - IGP, e - EGP, ? - incomplete VPN RD: 192.168.102.1:161 Network Next Hop Metric LocPrf Weight Path > 10.0.0.0/30 0.0.0.0 0 100 32768 ? > 10.255.255.252/30 0.0.0.0 0 100 32768 ? VPN RD: 192.168.102.1:210 Network Next Hop Metric LocPrf Weight Path > 10.98.50.1/32 0.0.0.0 0 100 32768 ? > 10.98.50.2/32 0.0.0.0 0 100 32768 ? >i 10.98.50.10/32 192.168.102.6 0 100 100 i > 11.0.56.0/24 0.0.0.0 0 100 32768 ? i 192.168.203.1/32 192.168.202.1 0 100 100 6871 i i 192.168.203.4/30 192.168.202.1 0 100 100 6871 i VPN RD: 192.168.102.1:211 Network Next Hop Metric LocPrf Weight Path > 10.98.50.1/32 0.0.0.0 0 100 32768 ? > 10.98.50.2/32 0.0.0.0 0 100 32768 ? >i 10.98.50.10/32 192.168.102.6 0 100 100 i --snip-- Has anyone else seen anything like this before? Is anyone successfully running 6.3 with BGP advertising the vpn address family in their network? Thanks Mark -------------- next part -------------- An HTML attachment was scrubbed... URL: