Great video, Nick. A small suggestion, please speak louder or raise the volume on the mic because, even with my PC volume set to max, your voice is still very low. Audible, but requires close attention.
Nice tutorial Nick! Thank you. I do have a question. What to do when there are multiple links to the same spine? Do we create a port-channel? And also how do we ensure proper load balancing? should we enable ecmp?
Hello, are you meaning multiple links from the same leaf to the same spine? It’s not common but that is possible yes. This would only be in the underlay and OSPF will take care of multi-path over those links and the BGP and overlay setup would be the same and no port-channels required, all standalone L3 links. Worth noting that if you have multiple links in this way, they MUST be the same speed and therefore the same OSPF cost. Hope this helps.
@nickcarlton4604 yes I'm thinking of a non-blocking architecture where there are 16hosts with 16ports as uplinks to the spines. So, if we are not doing port-channels then are you saying we should create multiple loopback interfaces per port and form ospf neighbors that way?
hi nick - many thanks for the videos - they're great - could you possibly share the actual ESXi setup & version and how you got the NX-OSv units on there - did you import the OVA's? if so could you provide info on the setup?
Hi I found this is using multicast in vxlan. Could you explain further the difference between unicast and multicast in vxlan? Currently our network is running on unicast worry there will be broadcast storm or flooding when we move to multicast network as we are not experienced.
Hi Nick. i'm trying to do the same with 2 nexus as VTEPs and few Arista as SPINEs. I'm having issues on carving the tcam region on the VTEPs nexus, to allow more arp-ether bandwidth. Every time I carve some space as you have done with the vpc convergence, i copy run start the config and then reload. After the reload is like I never done the carving, it doesn't save this modification. I am running NXOS: version 10.1(1), did you also encounter the same problem?
Nice tutorial. Thank you Nick!
this is a great video and awesome work you have done. thanks
Great video, Nick. A small suggestion, please speak louder or raise the volume on the mic because, even with my PC volume set to max, your voice is still very low. Audible, but requires close attention.
Thanks, I’ve adjusted my mic volume in subsequent videos
Great video. It would also be interesting to add on top of this the configuration of: 1) DHCP for one VLAN 2) VPC among two leaf switches.
Nice tutorial Nick! Thank you.
I do have a question. What to do when there are multiple links to the same spine? Do we create a port-channel? And also how do we ensure proper load balancing? should we enable ecmp?
Hello, are you meaning multiple links from the same leaf to the same spine? It’s not common but that is possible yes. This would only be in the underlay and OSPF will take care of multi-path over those links and the BGP and overlay setup would be the same and no port-channels required, all standalone L3 links. Worth noting that if you have multiple links in this way, they MUST be the same speed and therefore the same OSPF cost. Hope this helps.
@nickcarlton4604 yes I'm thinking of a non-blocking architecture where there are 16hosts with 16ports as uplinks to the spines. So, if we are not doing port-channels then are you saying we should create multiple loopback interfaces per port and form ospf neighbors that way?
Thanks Nick this has been great
thnks for sharing Nick.
Thanks, please explain how to monitor internet usage for each tenant
thank you
hi nick - many thanks for the videos - they're great - could you possibly share the actual ESXi setup & version and how you got the NX-OSv units on there - did you import the OVA's? if so could you provide info on the setup?
Hi I found this is using multicast in vxlan. Could you explain further the difference between unicast and multicast in vxlan? Currently our network is running on unicast worry there will be broadcast storm or flooding when we move to multicast network as we are not experienced.
Would you share your regex values for the syntax highlighting in iTerm? Your setup is one of the cleanest I’ve seen.
hi nick. many thanks for the video. were you doing it on real switches or simulator?
Hello, in this video and all videos featuring Nexus Switches. I’m using the NX-OSv Images from Cisco running in ESXI. Hope this helps!
Hi Nick. i'm trying to do the same with 2 nexus as VTEPs and few Arista as SPINEs. I'm having issues on carving the tcam region on the VTEPs nexus, to allow more arp-ether bandwidth. Every time I carve some space as you have done with the vpc convergence, i copy run start the config and then reload. After the reload is like I never done the carving, it doesn't save this modification. I am running NXOS: version 10.1(1), did you also encounter the same problem?
I solved changing the NXOS version