VXLAN's ACI2.1 # VXLAN ARCHITECTURE AND EXPLAINED IN COMPUTER NETWORK :
We saw that VxLAN created virtual Layer2 segments called VNI's (VNI = Bridge Domain). VNI run over the top of a layer-3 network. VxLAN switches use a special interface called a VTEP. This bridge VNIs to the layer3 network. When traffic comes in the VTEP encapsulates the traffic and sends it to a destination VTEP (UDP + IP + Message) where it is decapsulated.
# VxLAN Header :Of al these headers, the VxLAN header is the only one that's a bit different. All the other are well known. Fortunately, it's not difficult to decipher. There are four parts to the VxLAN header. 8 bits are reserved for future use. This is set to zero and ignored by the receiving VTEP. The VNI field is 24-bits long and contains the VxLAN ID. This large address space is what make it possible to have so many VNI's. Another 24-bits are reserved. As before, this field is ignored by the receiver. At the start of the frame is eight flag bits. Right now only bit 3 is used. Bit threee is the "i" field. Which is set to "1" for a valid VNI. The reset are seserved and ignore. So, there is a lot of unused space in the VxLAN header. It will be very interesting to see how this is used in future. If you have any throughts on these reserved fields, The extra VxLAN, UDP and IP headers add up to around 50 bytes of overhead. To account for this, you will need to enable jumbo frames everywhere. Otherwise you will get fragmentation which as know decreases performance. In the third part of this series, We're going to look at the spine leaf topology.
# Encapsulation :We start with an ordinary ethernet frame that a host would send. We call this the 'inner MAC frame'. This includes data, MAC address information, and other ethernet fields. It also may have a vlan tag included. In our example, traffice will stay within the VNI, so there is no routing required. The host sends the frame to the switch. The switch adds a VXLAN header, which contains the VNI . The VTEP now adds sveral additional headers, preservong the onner frame. VxLAN uses UDP for transport. The destination port is 4789 and the source port is random. ECMP, If available use a hashing algorithm to decide which link to put the traffice on. The random source port helps the algorithm to utilize the links evenly. An IP header is now added with the address of the destination VTEP. An ethernet header with a MAC address is added for delivery to the next physical device. As normal, the source and destination MAC addresses change with each device they pass through when the traffice arrives at the destination VTEP, the headers are removed, leaving the original frame, which can now be delivered to the host.
# VxLAN FRAME FORMAT :
You can see First format (Ethernet header + Payload + FCS) is defined as "End device transmit ethernet frames towards the upstream VxLAN switch".
Second format (VxLAN + Original Layer-2 frame) is defined as "The upstream switch attaches a VxLAN headers to the original frame. This header contains value 16,000,000 Layer-2 segments as opposite to the 4096 in 802.1Q".
Third format (UDP + VxLAN + Original Layer 2 frame) is defined as In addition, this is encapsulation in a User Datagram Protocol(UDP) segment.
Fourth format (Outer MAC + Outer IP + UDP + VxlAN + Original Layer-2 frame) is defined as next, an Outer IP & MAC header are applied to allow for VTEP tunneling instead a VLAN tag, Uses a VNID to distinguish Layer-2 segments.
# FULLY EXPLAINATION OF VxLAN :
We are trying to explain the VxLAN by picturizing with description. Hope you will better understand, Let's start First PIC (VLAN - Switch/Learning traffic) is defined as
I) WS-1 want to send traffic WS-3 in VLAN 11.
II) WS-1 initiates an ARP request to discover the MAC address of WS-3.
Let's start with Second PIC of (VLAN - Switch/Learning traffic) is defined as
III) SW-1 receives the request and floods the request out all ports except the one it was received on.
IV) SW-1 updates its own internal switching table with MAC address of WS-1
V) SW-2 forward the response and records the MAC address of WS-3.
VI) If WS-1 want to transmit with another VLAN then L3 Device has to route between (InterVlan).
Now, We will stat with VxLAN discussion with Third PIC (VxLAN-Switch/Learning traffic) is defined as
I) VxLAN doesn't use direct trunk connection for transmitted data.
II) VxLAN typically utilizes a Spine-n-Leaf topology, unlike the other Core-Aggression-Access network design.
III) VxLAN replaces directly connected physical trunks links with tunnels that provide similar functionality.
IV) VxLAN calls these multipoint tunnels VTEPs (VxLAN tunnel Endpoints).
Fourth PIC with (VxLAN-Switch/Learning traffic) is defined as
V) When setting up a GRE tunnel,You need to specify and destination addresses. For example, Loopback addresses on both switches are used and example uses GRE, Where you can use any interface for the tunnel source and destination. In VxLAN,loopback are the tunnel anchors.
VI) In order for each device to reach the loopback interface, an interior gatway routing protocol is needed.
Fifth PIC with (VxLAN-Switch/Learning traffic) is defined as
VII) The IP network being used for transport along with the IGP is reffered to as a UNDERLAY.
VIII) VxLAN use the VTEP interface as trunk and the VLAN ID is replaced with VNIDs to distiguish the Layer-2 segments.
Sixth PIC with (VxLAN-Switch/Learning traffic) is defined as
IX) The usual mechnisms for BUM traffic (broadcast, multicast and unknown unicast) are not available.
X) Since the transport network is IP-based, a multicast group is mapped to the L2 VNID to emulate the BUM traffice processing.
VxLAN-Switch/Learning traffic Analysis :
Seventh PIC with (VxLAN-Switch/Learning traffic Analysis) is defined as
I) WS-1 want to send traffice towards WS-3 in VLAN 11.
II) WS-1 initiates an ARP request to discover the MAC address of WS-3.
III) SW-1 receives the ARP request, adds the VxLAN tunnel headers and forwards it to the assigned multicast group.
VxLAN-Switch/Learning traffic Analysis :
Eighth PIC with (VxLAN-Switch/Learning traffic Analysis) is defined as
IV) The multicast group replicates the packets to all the VTEP that are part of the Layer-2 VNID.
VxLAN-Switch/Learning traffic Analysis :
Nineth PIC with (VxLAN-Switch/Learning traffic Analysis) is defined as
V) WS-3 receives the ARP request and replies to WS-1 by unicast.
VxLAN-Switch/Learning traffic Analysis :
Tenth PIC with (VxLAN-Switch/Learning traffic Analysis) is defined as
VI) Now all the traffice between WS-1 and WS-3 will flow directly through the overlay. This process is called VxLAN BRIDGING (traffice within the same L2 VNI)
VxLAN Routing :
Eleventh PIC with (VxLAN Routing) is defined as
I) Traffice following from WS-1 to WS-4 has to be routed via the L3 VNI to reach WS-4.
II) When traffic is destined to a different L2 VNI the process is called VxLAN routing. This used the L3 VNI and is similar to Inter-VLAN routing. The L3 VNI is mapped to a VRF.
Wrap Up :
- VxLAN supports 16,000,000 Layer 2 segments.
- VxLAN use s an IP-Based transport network.
- VLANs are commonly mapped to VxLAN L2VNIs
- Networking mechanics with VxLAN have some similarties to traditional 802.1Q VLAN mechanics.
- VxLAN traffice between can be bridged (L2, as within a VLAN) or routed (L3, as between VLANs).
Be updated into yourself and improve lives through DIT.
0 Comments