One of the main concerns when adding any new technology to your infrastructure is creating a more complex architecture.
While an SD-WAN can help simplify your overall network architecture and management, it is important to assess your network to eliminate as much complexity as possible while planning your migration.
The advantages of SD-WAN technology over a traditional network infrastructure can include flexibility in network architecture, seamless scalability, and significant cost savings. However, these benefits shouldn’t be hindered by the cost of an overly complex network. During the migration of an enterprise network, avoiding the introduction of additional complexity to the process becomes paramount.
Here are a few of the best practices to simplify the transformation of your organization’s network infrastructure to one using an SD-WAN approach. Follow this advice to ensure fast application performance and overall network security after the seamless migration is complete.
1. Perform a Thorough Network Assessment
Before formulating any migration plan, a thorough assessment of your network is a must. This includes examining the current view, as well as the expected network architecture after migrating to an SD-WAN model. After this process is complete, use this information to create a detailed plan outlining the entire migration.
A mapping document identifying the current network’s site types and their SD-WAN equivalents is an essential part of this dual assessment. Use network diagrams to analyze the current infrastructure, including information on costs per site and the current IP addressing model. Detailing the current bandwidth usage across the network is another must.
Reach out to network management personnel for additional details on any problems with the current setup, including application performance issues, security concerns, and equipment obsolescence. Remember, following the SD-WAN model lets organizations reduce their capital expenditures on on-premises equipment in lieu of an operational approach typical of cloud-based services.
Speaking of the cloud, any current initiatives on implementing service-based architectures for internal applications need to be part of this conversation. Company-wide security policies must be considered as well.
Ultimately, this planning and assessment effort remains an essential part of ensuring a seamless migration to an SD-WAN network infrastructure. In the end, this upfront work makes the process simpler.
2. Evaluate Functions That Can Move Out of the Branch
An SD-WAN networking approach lets your organization centralize the management of its entire network infrastructure. Remote management provides cost savings along with a subsequent boost in efficiency. While some applications and data might still require on-premises hosting–especially if cybersecurity or data governance is a concern– this becomes easier when using SD-WAN. In the case of security, consider a cloud access security broker (CASB) to both monitor activity and enforce the company’s security policies.
One of the main reasons cloud computing has become so popular is because organizations save money by offloading business functions to other companies specializing in that area. For example, if your company includes voice as part of its current network setup, consider moving that function to a VOIP provider. Engaging an IaaS or DBaaS provider for corporate databases is another smart idea using the same rationale.
Enterprises with a significant investment in MPLS transports also need to consider a hybrid architecture for their network infrastructure. Including both MPLS and SD-WAN offers high reliability while still providing enough flexibility and scaling for future growth.
In short, centralized network administration becomes simpler with an SD-WAN architecture. Imagine your company running remote branches with minimal (or no) IT personnel. Hosting business functions in the cloud, with automatic load balancing and monitoring from a central location, helps make it possible.
3. Assign IP Address Space to Make Network Management Easier
Some forethought when assigning the IP address space for your company’s SD-WAN network is necessary. Your network engineers must quickly be able to determine traffic origins, but you also need to consider the space needed for future growth. An approach where each branch gets its own unique address space makes sense, as it makes inter-branch communication easier.
Whatever the approach for IP addressing, try to ensure no overlapping addresses exist between branches. These kinds of conflicts prevent the use of a VPN. However, proper address segmentation helps mitigate the issue.
4. Choose the Best Server Configuration for Your Migration
Finally, server configuration needs to be considered. A Dynamic Host Configuration Protocol (DHCP) server makes sense for larger branches needing many addresses; configure your Customer-Premises Equipment (CPE) as a server in this scenario. Static (Internet Protocol) IP addresses are likely sufficient for smaller branches. Once again, watch out for overlapping address space.
When it comes to Domain Name Server (DNS), either a public server or a private, centralized server is the wisest choice. A platform-agnostic approach ensures consistency across the entire network.
Don’t forget about time synchronization. An orchestrator needs to understand the proper sequence of network events. Because of this, using a common Network Time Protocol (NTP) server is vital. A public option is fine, provided all branches are able to see the server. If not, hosting a private NTP server is a requirement.
Ultimately, migrating to an SD-WAN is a simple process–provided the right upfront planning takes place–no matter the complexity of the network architecture. For more information on how to simplify and optimize your company’s SD-WAN migration, check out our Enterprise Analyst white paper highlighting a few best practices.
One of the main concerns when adding any new technology to your infrastructure is creating a more complex architecture. 
