How Partial Failover Is Achieved With Layer 2 Networking

November 19, 2015 Disaster Recovery Blog, Networking Blog Articles

Most disaster recovery plans are all-or-nothing due to complicated networking. But what do you do when only one server or application fails – and other processes are still depending on it? That’s when partial failover comes into play, achieved by layer 2 networking.


What Is Layer 2 Networking?

Also known as layer 2 communications, layer 2 networking refers to the data link layer of computer networking. We’re about to get a little nerdy, so for the faint of heart, feel free to skip the next paragraph – just know that this networking layer involves the transfer of data to and from your computer.

According to the OSI model (also known as the Open Systems Interconnection model, but let’s be real, nobody calls it that), there are 7 layers of computer networking:

  1. The physical layer (your hardware) 
  2. Data link (layer 2 
  3. Network 
  4. Transport 
  5. Session 
  6. Presentation 
  7. Application 

As you can see, the OSI model is basically a sketch of how a network builds up from your hardware, all the way to the what you see on your computer screen (in this case, our gorg-eous website). The second layer, or data link layer, is the layer where data transfer takes place between your system, your network, and another system, all facilitated by a switch.


Partial Failover Explained

Partial failover is necessary when your primary infrastructure is still up and running, but only one (or a subset) of your servers, applications, or virtual machines (VMs) is experiencing problems. In this scenario, a complete site failover is unnecessary, but something still needs to be done about the corrupted system(s). Partial failover allows you to run the corrupted systems at your target (like a cloud host), while the rest of your functional systems keep running on your primary site.

For example: imagine that your email server fails. Just because you can’t send emails doesn’t mean you can let your entire company come to a halt (despite how much you’re enjoying the break from email-writing). With partial failover, you can specifically run your email server (or whichever server or application you choose) from your target, and have it back up and running within seconds – even during the busiest hours of your workday.

Partial failover is built out of layer 2 networking. This data link layer enables secure communication between your primary systems and your target in a unique way that gives you more than just an all-or-nothing failover solution. This means that your high availability solutions will be flexible for any situation; whether all of your systems fail, or just a subset, you can fail over to the target site of your choosing.


The Results of Using Layer 2 for Partial Failover

In a nutshell, the benefits of having partial failover under your belt include:

  • Ultimate failover flexibility for all types of disaster scenarios 
  • Interconnects multiple sites without changing operating system configurations 
  • Provides complete control over switching and routing between company branches 
  • Constant monitoring of replicated data for error detection and prevention in data sending 
  • Built-in Quality of Service (QoS), so latency-sensitive traffic (e.g. voice or video) is treated with preference over less sensitive traffic 

Sometimes tricky engineering makes for simplified solutions.

For more information about partial failover, contact us today at 317-707-3941.


Back to blog list