Over the last decade, data centres have quickly increased server virtualization, dramatically advancing agility and flexibility.
It is easier to monitor, automate, and orchestrate a network that has been virtualized and decoupled from the actual network.
The network overlays are the different virtual networks of interconnected nodes that enable applications to be deployed without having to change any of the underlying networks.
Virtual Extensible LAN (VXLAN) and Location/Identifier Separation Protocol (LISP) are two new encapsulation frame types designed specifically for data centres.
Key Takeaways
- Cisco LISP (Locator/ID Separation Protocol) is a network architecture that separates identity and location, while VXLAN (Virtual Extensible LAN) is a network overlay technology.
- LISP improves routing efficiency and scalability, while VXLAN enables the creation of large-scale, multi-tenant data center networks.
- Both LISP and VXLAN are used to address the limitations of traditional IP networks, but their approaches and use cases differ.
Cisco LISP vs VXLAN
The difference between Cisco LISP and Cisco VXLAN is that LISP stands for Locator Identifier Separation Protocol, and its main vendors are Cisco, VMware, Citrix, Red hat, Brocade, and Broadcom, whereas VXLAN stands for Virtual Extensible LAN, and its main vendors are Cisco.
LISP is the short form for Locator/Identifier Separation Protocol. It is a protocol that is meant for mapping and encapsulation. In a LISP environment, there are certain important environments.
These are LISP sites, LISP mapping sites, and non-LISP sites. It has a 24-bit LISP instance ID for overlay identification.
VXLAN is a technology that allows any IP routing protocol to be used to overlay a Layer 2 (L2) network over a Layer 3 (L3) underlay.
The L2 Segment ID field is extended to 24 bits in VXLAN, allowing up to 16 million separate L2 segments to exist on the same network. VXLAN encapsulates an L2 frame in an IP-UDP header, providing L2 connectivity across router boundaries.
Comparison Table
Parameters of Comparison | Cisco LISP | Cisco VXLAN |
---|---|---|
Full-Form | Locator/Identifier Separation Protocol | Virtual Extensible LAN |
Overlay Identification | 24-bit LISP instance ID | 24-bit virtual network ID (VNI) |
Fragmentation | Stateless and Stateful LISP methods. | MTU should be increased by 50 bytes to avoid fragmentation of VXLAN packets. |
Switch Support- Virtual | Cisco Nexus 1000v Virtual Switch | Cisco Nexus 1000v and VMware DVS |
Vendors | Cisco, VMware, Citrix, Red hat, Brocade, and Broadcom. | Cisco |
Support in Hardware | Arista 7150 and brocade ADX | No |
What is LISP?
Locator ID Separation Protocol is a type of network architecture that helps with the implementation of the use of 2 different name places instead of any single IP address. These two-name places are:
- EIDs: Also known as Endpoint identifiers, are assigned to different end hosts.
- RLOCs: Also known as Routing Locators, are assigned to the primary routers, I.e., devices that make up the global routing systems.
LISP sites are the named place where different EID exists. A LISP mapping service is an infrastructure that handles the EID to RLOC mapping for better efficiency. A non-LISP site is where the RLOCs are found.
There are two sites in a LISP site, namely, site 1 and site 2. Both of these sites have a host and a router. The hosts have an EID address, while the routers consist of an RLOC address.
LISP is a tunnelling protocol that employs a DNS-style mechanism to determine which router IP packets should be sent to.
The encapsulating and de-encapsulating LISP routers have a name:
- ITR (Ingress Tunnel Router): An ITR encapsulates IP packets.
- Egress Tunnel Router (ETR): De-encapsulates LISP-encapsulated IP packets.
- Tunnel Router (TR): A router that can handle both ITR and ETR traffic.
What is Cisco VXLAN?
VXLAN is a technology that allows you to segment your networks as VLANs do, but it also addresses VLANs’ scaling limitations and offers benefits that VLANs don’t.
Any IP routing protocol can be used to overlay a Layer 2 (L2) network over a Layer 3 (L3) underlay using VXLAN.
It employs the MAC-in-UDP encapsulation technique. VXLAN has the following benefits:
- Traditional VLANs offer 4K VNIs (broadcast domains), whereas 16M VNIs (broadcast domains) are available here.
- This functionality allows L2 to be extended anywhere in an IP network.
- You can allocate the resources in a more appropriate way among the different centres.
- Flooding has been optimized.
- They provide network segmentation that supports large numbers of tenants that are required by cloud builders.
- It enables you to transfer the virtual machines across the servers incorporated into the Layer 2 domains with the help of the tunnelling of traffic that is done over the Layer 3 networks.
Layer 2 Ethernet frames are embedded in Layer 3 UDP packets using VXLAN tunnelling technology, allowing you to build virtualized Layer 2 subnets or portions that span physical Layer 3 networks.
VXLAN network identification (VNI) is assigned to each Layer 2 subnet to segment traffic.
Main Differences Cisco LISP and VXLAN
- Cisco LISP stands for Locator/Identifier Separation Protocol, and Cisco VXLAN stands for Virtual Extensible LAN.
- LISP has a 24-bit LISP instance ID for overlay identification, whereas VXLAN uses a 24-bit virtual network ID for overlay identification.
- For the fragmentation processes, LISP uses stateless and stateful LISP methods. On the other hand, for VXLAN, MTU should be increased by 50 bytes to avoid fragmentation of the VXLAN packets.
- LISP has a Cisco Nexus 1000v Virtual Switch for the various virtual switch supports, whereas VXLAN uses a Cisco Nexus 1000v and VMware DVS.
- LISP has Cisco, VMware, Broadcom, Brocade, Red Hat, and Citric as its main vendors, and VXLAN only has Cisco as its main vendor.
- The support for hardware offered by LISP is available for Arista 7150 and Brocade ADX, whereas VXLAN does not support any support for hardware.