vIST and Extreme Networks

_ 04/02/2025_ Yoann Jonard

vIST and Extreme Networks

How does vIST clustering mechanism work and what are its advantages ?

Introduction

Network reliability is now critical, as businesses cannot afford downtime, users do not tolerate downtime, and network administrators face challenges to find maintenance windows to perform disruptive tasks on the network infrastructure. We could keep working before business hours, or after, or even on Saturday and Sunday, or we could leverage reliable technology to assure a network resilience we can trust, and we can use.

Traditional network designs often relied on Spanning Tree Protocol (STP) to prevent loops, but this came at the cost of unused redundant links. Split MultiLink Trunking (SMLT) improved upon this by allowing the use of redundant links between the access layer switch and the distribution switches with load balancing while providing subsecond failover in case of switch failure.

However, how can one have multiple links coming from the same switch and going to 2 different switches, while keeping these links active ? That’s where vIST comes into place.

What is Virtual Inter-Switch Trunk ?

vIST is a new clustering mechanism enhancing traditionnal IST where a physical link (layer 1) was mandatory between the 2 switches being clustered. Thanks to the SPBM Fabric, this requirement is no longer required, and the vIST will create a virtualized channel through the Shortest Path Bridging (SPB) Fabric. This virtual channel connects two devices, forming a virtualized Switch Cluster. As long as there is a path within the Fabric for the vIST information to flow through, the cluster will stay up and running, giving you maximum flexibility while building your clusters. The beauty of vIST lies in its ability to carry both control traffic and data traffic during a failover, dramatically improving resiliency over other methods.

Let’s have a look at a short video showing how reliable is a cluster built with vIST.

Key Benefits of vIST

Elimination of Single Points of Failure: Unlike traditional IST implementations that use a dedicated MultiLink Trunk (MLT), vIST eliminates potential single points of failure. The virtual channel remains operational as long as there’s SPBM connectivity between the vIST peers.
Interoperability: vIST can operate between any two devices that support it, regardless of their type. This flexibility allows for easier integration in diverse network environments.
Simplified Configuration: With Simplified vIST, network administrators can enable virtual IST for SMLT deployments using a single CLI command, streamlining the setup process. This simplified process comes with the cost of both switches not running SPBM. The simplified vIST architecture is suited for really specific environments.

Implementation Considerations

While vIST offers significant advantages, there are some important considerations for implementation:

The system MTU must be at least 1950 bytes due to increased header size when transmitting packets over the SPBM cloud.
For proper traffic flow, Layer 2 VSNs created on one vIST peer must also be created on the other.
Updates to vIST VLAN IP addresses should be performed during maintenance hours to prevent data loss.

Required configurations

Below is an example configuration for a cluster using VLAN 10, I-SID 1010 and subnet 192.168.1.0/30.

As only 2 IPs are needed, we can use a /30. This ensures that only these 2 IPs can exist, and that we don’t consume a larger subnet unnecessarily.

It is highly recommended that you use a different I-SID per cluster, but you can use the same VLAN ID for all your clusters. In the Fabric, the I-SID is decisive, not the VLAN ID. You can therefore use VLAN 10 for all your clusters, using I-SIDs 1010, 1011, 1012, etc.

switch 1:
  vlan create 10 type port-mstprstp 1
  vlan i-sid 10 1010
  interface vlan 10
  ip address 192.168.1.1/30
  exit
  virtual-ist peer-ip 192.168.1.2 vlan 10

switch 2:
  vlan create 10 type port-mstprstp 1
  vlan i-sid 10 1010
  interface vlan 10
  ip address 192.168.1.2/30
  exit
  virtual-ist peer-ip 192.168.1.1 vlan 10

The Future of Network Resilience

As networks continue to grow in complexity and importance, technologies like vIST will play a crucial role in ensuring uninterrupted service. By providing robust failover mechanisms and eliminating single points of failure, vIST represents a significant step forward in network design and management. For network administrators looking to enhance their infrastructure’s resilience, vIST offers a powerful solution that combines the benefits of clustering with the flexibility and reliability of virtualized networking.