A recent SolarWinds survey on IPv6 transition revealed that two-thirds of respondents comprising of network administrators and engineers are considering, planning to, or already have migrated from IPv4 to IPv6. With the migration trend going up, another notable indication was that 47% said they are “not at all confident” that their company has an actionable IPv6 adoption plan in place.

 

 

The aforementioned statistic is a big red flag and a call for serious concern.  It’s imperative that network administrators are fully educated on the various IPv6 transition techniques and the factors that could potentially impact the enterprise network because not knowing could be detrimental. 


So let’s dive in to understand how a dual-stack architecture that utilizes the Dual Stack Transition Mechanism (DSTM) can be a feasible approach to IPv6 migration.


What is a Dual IP Stack?


Dual-stack is one of the most widely adopted techniques for IPv6 migration. It helps to establish communication between your IPv6 network and the native IPv4 hosts and applications. A dual-stack node has support for both protocol versions and is referred to as an IPv6/IPv4 node. IPv6/IPv4 nodes may have a ‘configuration switch’ to enable or disable one of the stacks, which means they can have three modes of operation:


  • IPv4 only - IPv4 stack enabled and IPv6 stack disabled
  • IPv6 only - IPv6 stack enabled and IPv4 stack disabled
  • Both IPv4 and IPv6 stacks enabled

 

Dual Stack Transition Mechanism


Dual Stack Transition Mechanism (DSTM) is a transition mechanism based on the usage of IPv4-over-IPv6 tunnels to facilitate interoperability between newly deployed IPv6 networks and existing IPv4 networks. DSTM is best-suited for IPv6 dominant environments where hosts still need to exchange information with legacy IPv4 hosts or applications.


Significant Advantages:

  • Transparent to the network and to the application
  • Legacy IPv4 applications can be run over IPv6-only networks without modification
  • IPv4 addresses are dynamically allocated as needed and then reclaimed
  • Based on standard protocols

 

How does DSTM work?


The DSTM architecture consists of three major parts.

  • DSTM Client - dual-stack node running the DSTM client software which requests the IPv4 address from the DSTM server; allows host on the IPv6 network to communicate with IPv4 applications or hosts.
  • DSTM Server - a dual-stack node running the DSTM server software which provides IPv4 address allocation along with the IPv6 address of the DSTM Gateway / TEP.
  • DSTM Gateway or Tunnel End Point (TEP) - performs the encapsulation and decapsulation of tunneled packets.

 

When a host in the IPv6-only domain needs to communicate in IPv4, it queries the DSTM server for a temporary IPv4 address. The DSTM server provides a temporary IPv4 address for the host from the address pool, including its validity time, along with the IPv6 TEP address. Following this, the host (DSTM client) builds its IPv4 packet with the allocated address information and forwards to the TEP. The TEP then decapsulates the packet and forwards it to the destination. All the IPv4 packets coming from the client are tunneled to the TEP to perform encapsulation and decapsulation of the IPv4 packets. The TEP stores the mappings between the IPv4 and IPv6 addresses.


Conclusion


It's crucial that network administrators fully understand the different approaches to IPv6 to know which one is right for their network. The Dual Stack Transition Mechanism (DSTM) can be an ideal approach for early adopters of IPv6 with an IPv6 dominant environment that still need to communicate with legacy IPv4 nodes.


Regardless of the migration strategy chosen, the most important aspect is planning ahead. Migration to IPv6 doesn’t have to be complex as long as you understand your requirements and have a well-thought-out plan. Even better, you don’t have to go it alone. Equipping yourself with the right IP address management tool can help ensure a smooth transition and provide easy ongoing management of your dual-stack network.