God bless you, Mr. Gould. This was driving me crazy. I only have a CCNA, and I'm trying to work with the CSR1000v in my CML and had no clue what the hell a service instance was/does. Your explanation of "rewrite ingress tag pop 1 symmetric" made it all click for me. I was sending double tagged traffic from my switch to the router. Saved me hours of heartache. Subscribed for more. Wishing you all of the success in the world, mon ami!
You're welcome "ANCAP Prince"! I'm so glad the video was helpful.... i have more videos on CML... ruclips.net/video/C7gA2a64yBI/видео.html ruclips.net/video/1td1FDL0-Iw/видео.html ruclips.net/video/rkb1W03ibBo/видео.html ruclips.net/video/SrryHbjpnAc/видео.html ruclips.net/video/pHb9sC-k5Ss/видео.html
Hi Aaron ...I have been following your videos for quite sometime now. Its wonderful. Please make a video on QinQ over EVPN MPLS with c-vlan and p-vlan. Thanks
tnx nicely explained. is this bridge-domain command simillar to juniper ? if you have one more interface in BD100, it would be part of same broadcast domain ? and you could reach outside of that broadcast domain, through the BDI100 interface (that is, to route) ? edit:actually, creating efp logical interface on the physical interface and puting it in bridge domain is the same as putting logical unit in BD on juniper ? tnx edit2: actually, that config on csr1000 is the same thing if you would have juniper router with vlan-tagged unit, and irb routing interface which are member of bridge domain : D ?
yw! yes there are similarities between cisco and juniper and other vendors for that matter. they all have their own CLI syntax of course, but bridging and routing concepts are the same. it's just a matter of how to accomplish it per vendor.
@@aarontechtalk oh, and 12:15 ,.so this example, vs same case if you would make regular 0/1.100 subinterface on csr router, is different in what ? in that you can have per-port vlan meaning ? and in first case you have L2 port bridging ethernet frames, and with subinterface its a L3 port ? am i right . tnx
I'm referring to the way ethernet switches operate, and how Cisco has a VTP Domain where vlans are advertised throughout the domain. So, you have this single vlan identity known everywhere in your traditional, enterprise or campus style network. So, as I compare that with the new MEF (Metro Ethernet Forum) and SP (Service Provider) style bridging, what you now have is a way of still accomplishing bridging, but, you can remove the vlan tag at the customer interface (UNI) and then bridge the traffic across the SP core, and then add a tag back on the traffic as you forward it to the far side customer interface (other UNI)... this is a pop and push operation upon ingress and subsequently upon egress. See what I mean?
yes absolutely, the original customer traffic IP header is definitely unchanged. But this should be obvious during bridging. It's fine, it's a good question. More to the point... bridging implies, forwarding based on the Layer 2 Ethernet header, which means the IP header isn't touched.... but, the point is, the 802.1q vlan tag can be popped... then the vanilla (untagged) ethernet header is bridged across the SP core.... and then you can decide to push/impose an 802.1q vlan tag upon egress bridging on the far-side CE handoff. I really should do a video to show and explain this. It's really good to know and understand. Hopefully I can knock that out asap.
Thank you for your explanation on this video...but i have still confused about which different efp and evc...from this video i got the point that the service instance command is efp it self...so how about evc? because previously I thought that the service instance was evc...thanks you sir...i hope you want answer my question
The definition of an EVC is the entire connection which includes 2 or more endpoints. So the endpoint is like the EFP (Ethernet Flow Point) If you have a point to point circuit, you will have 2 EFP's, but 1 EVC (Ethernet Virtual Connection). The EVC is the entire circuit connecting both endpoints. If you have a multipoint to multipoint circuit, with for example, 4 endpoints, you will have 4 EFP's...one for each endpoint, but 1 EVC (Ethernet Virtual Connection). The EVC is the entire circuit connecting all of the 4 endpoints. See what I mean?
I'm not demonstrating anything about the transport protocol here in this video, nor am I speaking of L3VPN. Here I'm only talking about an ethernet interface and how to send and receive frames with various tagging configurations.
ah, of course... many ways. If you need to emulate layer 2 between the 2 sites, you could either do it the old fashion way with switches in between and vlans... or the newer way, would be various Layer 2 VPN types....mef calls them eline, elan, etree... but the industry implements them using various sp core network virtualization methods, using mpls or segment routing... vpls, pseudowire using ldp or sr...with or without mpls. i understand that pbb can do it, but i've never used it. vxlan can also do this l2 emulation over a pure ip core. if Layer 3 is needed between the sites, then it's layer 3 vpn methods using bgp-based vpn's or ldp based.
God bless you, Mr. Gould. This was driving me crazy. I only have a CCNA, and I'm trying to work with the CSR1000v in my CML and had no clue what the hell a service instance was/does. Your explanation of "rewrite ingress tag pop 1 symmetric" made it all click for me. I was sending double tagged traffic from my switch to the router. Saved me hours of heartache. Subscribed for more. Wishing you all of the success in the world, mon ami!
You're welcome "ANCAP Prince"! I'm so glad the video was helpful.... i have more videos on CML...
ruclips.net/video/C7gA2a64yBI/видео.html
ruclips.net/video/1td1FDL0-Iw/видео.html
ruclips.net/video/rkb1W03ibBo/видео.html
ruclips.net/video/SrryHbjpnAc/видео.html
ruclips.net/video/pHb9sC-k5Ss/видео.html
Thanks, Aaron, very nice and clear
Thanks Ashraf, I’m glad it was helpful.
thank you, this is a really clear explanation on EFP
Thanks Tien, I'm glad it was helpful.
Great video, cleared up a lot for me
Great to hear it Evan!
Great video, nice and clear, help me a lot. Thanks,
Thanks David, I'm glad it helped.
Hi Aaron ...I have been following your videos for quite sometime now. Its wonderful.
Please make a video on QinQ over EVPN MPLS with c-vlan and p-vlan.
Thanks
Thanks Kuldeep, I'm glad you are enjoying the channel. I will try to make QinQ over EVPN MPLS video soon.
tnx nicely explained. is this bridge-domain command simillar to juniper ? if you have one more interface in BD100, it would be part of same broadcast domain ? and you could reach outside of that broadcast domain, through the BDI100 interface (that is, to route) ? edit:actually, creating efp logical interface on the physical interface and puting it in bridge domain is the same as putting logical unit in BD on juniper ? tnx
edit2: actually, that config on csr1000 is the same thing if you would have juniper router with vlan-tagged unit, and irb routing interface which are member of bridge domain : D ?
yw! yes there are similarities between cisco and juniper and other vendors for that matter. they all have their own CLI syntax of course, but bridging and routing concepts are the same. it's just a matter of how to accomplish it per vendor.
@@aarontechtalk oh, and 12:15 ,.so this example, vs same case if you would make regular 0/1.100 subinterface on csr router, is different in what ? in that you can have per-port vlan meaning ? and in first case you have L2 port bridging ethernet frames, and with subinterface its a L3 port ? am i right . tnx
I'm referring to the way ethernet switches operate, and how Cisco has a VTP Domain where vlans are advertised throughout the domain. So, you have this single vlan identity known everywhere in your traditional, enterprise or campus style network. So, as I compare that with the new MEF (Metro Ethernet Forum) and SP (Service Provider) style bridging, what you now have is a way of still accomplishing bridging, but, you can remove the vlan tag at the customer interface (UNI) and then bridge the traffic across the SP core, and then add a tag back on the traffic as you forward it to the far side customer interface (other UNI)... this is a pop and push operation upon ingress and subsequently upon egress. See what I mean?
@@aarontechtalk i see. by brdige the traffic across SP core, you mean its original ip header stays intact ?
yes absolutely, the original customer traffic IP header is definitely unchanged. But this should be obvious during bridging. It's fine, it's a good question. More to the point... bridging implies, forwarding based on the Layer 2 Ethernet header, which means the IP header isn't touched.... but, the point is, the 802.1q vlan tag can be popped... then the vanilla (untagged) ethernet header is bridged across the SP core.... and then you can decide to push/impose an 802.1q vlan tag upon egress bridging on the far-side CE handoff. I really should do a video to show and explain this. It's really good to know and understand. Hopefully I can knock that out asap.
Thank you for your explanation on this video...but i have still confused about which different efp and evc...from this video i got the point that the service instance command is efp it self...so how about evc? because previously I thought that the service instance was evc...thanks you sir...i hope you want answer my question
The definition of an EVC is the entire connection which includes 2 or more endpoints.
So the endpoint is like the EFP (Ethernet Flow Point)
If you have a point to point circuit, you will have 2 EFP's, but 1 EVC (Ethernet Virtual Connection). The EVC is the entire circuit connecting both endpoints.
If you have a multipoint to multipoint circuit, with for example, 4 endpoints, you will have 4 EFP's...one for each endpoint, but 1 EVC (Ethernet Virtual Connection). The EVC is the entire circuit connecting all of the 4 endpoints.
See what I mean?
tranks
yw
What woıld be actual transport protocol. L3VPN i guess?
I'm not demonstrating anything about the transport protocol here in this video, nor am I speaking of L3VPN. Here I'm only talking about an ethernet interface and how to send and receive frames with various tagging configurations.
@@aarontechtalk ok but how you achieve this in real world with 2 different locations? What will encapsulate the ethernet frame?
ah, of course... many ways. If you need to emulate layer 2 between the 2 sites, you could either do it the old fashion way with switches in between and vlans... or the newer way, would be various Layer 2 VPN types....mef calls them eline, elan, etree... but the industry implements them using various sp core network virtualization methods, using mpls or segment routing... vpls, pseudowire using ldp or sr...with or without mpls. i understand that pbb can do it, but i've never used it. vxlan can also do this l2 emulation over a pure ip core. if Layer 3 is needed between the sites, then it's layer 3 vpn methods using bgp-based vpn's or ldp based.
@@aarontechtalk thanks
This the answer that i need