What is an NSX-T Edge? BRIEF explanation!

[@aritrade5266]

Crisp and clear explanation. Great work Mike

[@Stingray7423]

The encouragement part, in the beginning, is as valuable as the rest of the video! Thank you!

[@majedalshehri2367]

I Can't thank you enough. NSX-T edge node topic was very challenging for me, but your 6 minute video cleared everything.

[@TechAndWanderlust]

I love watching this video like my grandpa always said you gotta put it where the goats can get it. Nice and easy clear concise to the point this video help me kill it in an interview.

[@ekatanaka]

Great explanation. I like the simple way that you explain it. Thanks.

[@badge2204]

this like is for your early morning !

[@laszlo.laszlo]

You are very good! Thank you very much. I'm very new in NSX but I like it.

[@NRDYTech]

thank you!

[@johnclark5172]

Thanks for sharing your knowledge, i really appreciate it!

[@MrManchile]

Good explanation. Also I like the bonus lecture - find time to do what is important to you.

[@sammirmajlesi6329]

Very nice videos. Excellent job. Thanks.

[@edwinmatos9435]

Well done! Good material.

[@williamrivera724]

Thanks for all your videos 👍☕️

[@umairriaz1275]

Thanks for sharing, really love your videos. great insight into world of nsx-t. please do videos on how does DNS, NAT, etc services work in NSX-T and in what scenario do we use these services from NSX-T.

[@NRDYTech]

Thanks Umair! I will start thinking about a way I can do those vids!

[@NRDYTech]

Update! I made a video addressing DNS! Check it out: kzbin.info/www/bejne/rp7Uo6F5eMmZfsU

[@Nafay1991]

well explained... need more knowledge on this

[@GG-kc6ie]

Great summary

[@amityadav83]

This is what I was looking for .... Enjoy best output Goto settings >speed 1.5x ..... enjoy.

[@michaelhall8039]

Thanks again!

[@cristhiansaid]

thks for your excelent video! i really appreciate!, I have a question, what about if i only have one cluster for all? i do not have a cluster for managment and another for production? is there any problem if i only manage and deploy into 1 cluster?

[@NRDYTech]

You can do that. Probably not recommended/best practice as a problem in your data plane (your edges & the host they're sitting on) could theoretically affect management of your environment (since it's on the same host(s)/cluster)- but technically it will work just fine.

[@jogandomais]

Awesome video, thank you

[@MrPankajkr06]

You are awesome 👍

[@SalsaBachataSingapore]

Nice video thanks . Does esxi host hosting edge node also need to have nvds like transport node ?

[@NRDYTech]

Ashish, the esxi host hosting the edge node does NOT need to be prepped for NSX-T at all, and it does *not* need an NVDS. The reason is because the edge node VM itself has an N-VDS inside of it :) Hope this helps.

[@SalsaBachataSingapore]

@@NRDYTech yes thanks a lot

[@fisherman18888]

I'm an ACI expert but still learning NSX. The hardest part of the design/architecture learning has been, how do edge nodes scale. I understand the sizes of VMs and why VM vs bare metal but nothing explains how you know when to size 2, 3, or more edge nodes, or even when to have more than one edge cluster and how many nodes might be in each. Any reading you could recommend on this type of scaling?

[@NRDYTech]

Hi JC! Glad to have you here. I worked with ACI in the 1.0/2.0 days, been a while. The first place I'd start is the design guide - communities.vmware.com/t5/VMware-NSX-Documents/VMware-NSX-T-Reference-Design/ta-p/2778093 - that said, it doesn't completely address your question. "When" to scale is a tough question to answer, because ultimately the answer is something like: when the resources of the Edge VM (CPU/Mem) OR the vSphere host it sits on - OR - the network connected to the vsphere host - is saturated, it's time to talk about scale. Obviously if the network is the bottleneck, we're talking about either adding bandwidth to the host, or additional edge nodes. Lots of different scenarios! My $.02 - most customers start with pretty stout edge VM form factors, and keep an eye on resource utilization. If they see they're coming up on limits, they deploy an additional edge VM, and swap it out with the "smaller" one, non-disruptively.

[@NRDYTech]

Keep in mind, if you're doing Active/Standby services for your T0 or T1, each respective T0 or T1 will only occupy 2x edge nodes at any given time..so having 8 edge node VM's may not make sense in that scenario. Often times what I see is large edge clusters when the customer is doing Active/Active on their T0. 90% of deployments will just be 2x edge VMs. Hopefully this helps a tiny bit

[@fisherman18888]

@@NRDYTech Much appreciated. Do you have a blog or contact outside of here? I'd pay a consulting fee or such to pick your brain on a few topics on an hour WebEx.