How VRF Lite works in NSX-T 3.1 (Pt. 1)

[@ahpek199]

great explanation with diagram. because out there many diagram is complex and confusing.

[@vgianni1]

Keep it up like this , awesome videos!

[@semyonbychkov9513]

Awesome videos, great!

[@craigjenkins2085]

great stuff thanks mike

[@vibskrusher]

Nice explanation Mike :) I have watched your udemy videos as well they are super cool.

[@NRDYTech]

Thank you!

[@ZafarKhan-th8wb]

Thank you so much for this explaination...cleared a lot of things for me

[@mehrandehghan884]

May i ask you to make a video about EVPN?

[@sliddjur]

How does routing between tenants work with BGP? You said the VRF will inherit BGP as number from t0? Now, in BGP, the first rule is that you have to discard routes with your own BGP ASN in the route?!

[@hamzaouladbenbrahim1615]

hello Mike, Hope you're doing well. I found out that one edge cluster can backup multiple T0

[@mandarverlekar7792]

Yes, you are right. However, when you try to assign uplink interfaces for the second T0 from the edges of the same edgecluster, it will throw an error saying that the interface is already assigned

[@anilanilkumar7495]

Excellent... in T0 we have two Different VRF Red & Blue can we communicate those VRF from T0 GW ? just some curiosity i am asking..

[@mehrandehghan884]

It was answered in minute 8:50

[@sliddjur]

Your videos are in the wrong order in the playlist :)

[@AEONMata]

Hi, it is really necessary to create another T0 edge cluster for tenants with old way? What about tenants on T1 router level,anyway thats how vCD works right?

[@NRDYTech]

Sure, you could do that as well, I suppose it depends on your requirements/preference. Most of my customers seem to prefer splitting up to the T0, but both are valid options in my opinion

[@pinpinpoola]

Only 2 'That said" in 10 minutes. :-)

[@NRDYTech]

I'm trying REALLY hard :)

[@johnosborne1977]

Annoyingly as I found out recently, not supported in NSX Federation (as well as LB, IDS/IPS, L7 Context FW). OK, to be fair there is nothing stopping someone creating VRFs using the local managers, but if you are using Federation I guess you would want a VRF to be subject to what Federation offers in terms of multi-site capability.

[@NRDYTech]

Yep! There's a few things on that list of unsupported federation features (great timing btw, I'm literally working in my lab on federation at this moment). LB is supported with federation though, albeit limited to only a T1 one-arm design. With 3.1 being the first full federation release, I'm sure a lot of that will be cleared up in the next major release.

[@johnosborne1977]

@@NRDYTech Apologies, the LB you are correct, I think its listed as not supporting L4+ Services (brain is on a go slow today). However the the doc where I got the limitation list from (NSX-T 3.1 Multi-locations Design Guide), I did find an error (and reported back into VMware) - it was stated that NSX management cluster did not support SRM for Disaster recovery, and in fact it has been since 3.0.2. I Think as these docs are so huge they have been written over a long period of time, its easy to miss making changes where a feature changes just prior to release. Anyway, good luck with the labbing, I am about to start testing VCD in my lab, and also seeing if I have enough resource to try VCF (nested version of).

[@NRDYTech]

@@johnosborne1977 Nice! Yes, I know one of the authors of that doc - and he's amazingly sharp, so I suspect you're right - that doc is HUGE and probably just fell through the cracks. Federation is going smooth - doing packet walks and such to prepare for a fed video on this channel. For the VCF lab - good luck! Best advice I have is common sense - remove mem/cpu reservations on everything, and try really hard to be stingy with allocating mem/cpu :) My lab is on the edge with federation - around 90% used, and VCF is worse! New 128GB host on its way.