Thank you for including the wan... no other video shows where the wan connection is in spine-leaf. The bit I don't understand is when it's bigger than a small datacentre... if you have 40 spines your rack switches would need 40 ports just to reach the spines. But 4 spines makes perfect sense just connect your top of rack switches to the 4 spines each
How will you build campus network using Spine and leaf as there will be too many cables, If there are 10 buildings it can be nightmare. How does this work in such a scenario ? Kindly guide me if you have any video on that.
Hi I would like to ask. Can this architecture be used for a network campus? and if so, where would devices such as laptop be placed? would it be attached to one of the switches? thk in advance :)
@@networkingwithh4716 oh i see, thk for answering my qns 😄 im trying to make a campus network but i’ll stick with a 2 tier topology. That video u made on 2 tier topology was very good btw, understood it very easily 😁
They are not same. One of the major difference i see is you can use layer 3 routing between spine and leaf in spine and leaf architecture. Any traffic between two nodes is just one hop away and it uses full available bandwidth. where as collapsed core or 2-tier hierarchical model is limited to spanning tree and vPC. which might not allow us to use full available bandwidth.
One of the major difference i see is you can use layer 3 routing between spine and leaf architecture. Any traffic between two nodes is just one hop away and it uses full available bandwidth. where as collapsed core or 2-tier hierarchical model is limited to spanning tree and vPC. which might not allow us to use full available bandwidth.
Hi, can you explain a bit about the non -blocking fabric. If there is no interconnection between leafs and spines themselves will create a non-blocking switch ? how ?
That is a good question... Non-blocking means that each port is capable of sending and receiving traffic at the maximum speed to and from any other port
Correct me if im wrong (and i dont think i am) You only have layer3 control beyond the "Spine" [Id still call them cores]. This can be depending on your application a bastard, and it can become a really big mess really fast on networks with vlans and lans within lans. So bear that in mind, also any network you have to throw hardware at to solve Latency means you did a bad job in the first instance and now youll spend the next 10 years chasing it. Any network that has zero scale options is built in from the start is a failure. Customer asks us for say a 10 gig network. Well ensure key switches have 20/40G and ensure every single switch has at least 4 free ports, and where those ports are free LCAP and bonding is available.
Best explanation of spine-leaf I've seen. I was having trouble understanding but now it makes sense...thank you!
Quick and concise explanation. Thank you!
Omg you actually made it so easy to understand! Most network engineers skip the obvious and assume everyone is an expert. Bravo.
Great :)
Thank you for including the wan... no other video shows where the wan connection is in spine-leaf. The bit I don't understand is when it's bigger than a small datacentre... if you have 40 spines your rack switches would need 40 ports just to reach the spines. But 4 spines makes perfect sense just connect your top of rack switches to the 4 spines each
Quick and concise explanation. Thank you!👍🏻
Very clear explanation
Glad you liked it
Simple and Clear.
Glad it helped
Excellent video. Thank you!
Glad you enjoyed it!
Excellent breakdown, to the point without adding alot of buzz words.
Thanks
Superb vid, thank you!
Thank you sir, well explained and easy to follow.
You are welcome
Good explanation and really skillful use of graphics and icons!
Glad it helped!
Thanks for sharing knowledge!
My pleasure
Thank you. Great video
Really good video, easy to understand!
Thanks 🙏🏻
Wow nice, easy and clear explanation Thanks for that keep it up
You are most welcome
Kept it simple and good explanation
Very well explained
Thanks Hassan
Great presentation. Have an upvote.
Thanks
Superbly Explained, Short and very informative. Thanks so much🙏
Love it.
Nicely explained 👍
Very nice explained
Thank you so much 🙂
Nice job
Thanks 😊
Obrigado!!!
Good explanation can you pls put more videos about ACI its very very useful for us.. marvelous explanation
thks!!
How will you build campus network using Spine and leaf as there will be too many cables, If there are 10 buildings it can be nightmare. How does this work in such a scenario ? Kindly guide me if you have any video on that.
Hi, EoR stands for End of Row not End of Rack.
Made it simpler!
Hi I would like to ask. Can this architecture be used for a network campus? and if so, where would devices such as laptop be placed? would it be attached to one of the switches? thk in advance :)
It is a nice question... Spine and leaf are mostly used by data center networks. laptops are connected to Access network through Wi-Fi or lan Switch
@@networkingwithh4716 oh i see, thk for answering my qns 😄 im trying to make a campus network but i’ll stick with a 2 tier topology. That video u made on 2 tier topology was very good btw, understood it very easily 😁
Can you talk more about the concept of top or rack switch?
noted..
how full mesh ?
Hi sir, is two-tier architecture and spine-leaf architecture same?
They are not same.
One of the major difference i see is you can use layer 3 routing between spine and leaf in spine and leaf architecture.
Any traffic between two nodes is just one hop away and it uses full available bandwidth.
where as collapsed core or 2-tier hierarchical model is limited to spanning tree and vPC. which might not allow us to use full available bandwidth.
@@networkingwithh4716 Thank you sir, what about the cost which one is cheap?
@@kayazar5387 cost depends on what devices you choose to use..modern networks specially data centers prefer leaf and spine.
Hi sir, noob question, what is the difference between a collapse core and a spine-leaf?
One of the major difference i see is you can use layer 3 routing between spine and leaf architecture.
Any traffic between two nodes is just one hop away and it uses full available bandwidth.
where as collapsed core or 2-tier hierarchical model is limited to spanning tree and vPC. which might not allow us to use full available bandwidth.
Hi, can you explain a bit about the non -blocking fabric. If there is no interconnection between leafs and spines themselves will create a non-blocking switch ? how ?
That is a good question...
Non-blocking means that each port is capable of sending and receiving traffic at the maximum speed to and from any other port
Correct me if im wrong (and i dont think i am) You only have layer3 control beyond the "Spine" [Id still call them cores]. This can be depending on your application a bastard, and it can become a really big mess really fast on networks with vlans and lans within lans. So bear that in mind, also any network you have to throw hardware at to solve Latency means you did a bad job in the first instance and now youll spend the next 10 years chasing it. Any network that has zero scale options is built in from the start is a failure.
Customer asks us for say a 10 gig network.
Well ensure key switches have 20/40G and ensure every single switch has at least 4 free ports, and where those ports are free LCAP and bonding is available.