How to maintain network continuity in a DR strategy
Disaster doesn't just come in the form of fire, flood and ransomware. Loss of network continuity is a real issue and must be appropriately addressed in disaster recovery planning.
With an ever more interconnected world, reliable network connectivity is critical. It is also something few companies can exert control over.
SaaS tools, email, electronic data interchange, cloud providers and websites all require connectivity. A business without network continuity is open to serious financial loss and reputational damage. Information is the lifeblood of any modern organization, but network infrastructure can be fragile.
Network disasters take many forms, including equipment failure, power loss and human error. Network outages are further compounded by the fact that repairing the infrastructure takes time.
Network continuity is tough, not impossible
Fortunately, the increased dependence on networks hasn't gone unnoticed. Most modern routers have backup connectivity modules installed by default or as an optional add-on.
Thanks to the widespread adoption of 5G mobile networks, these devices can provide seamless coverage across backup connectivity to enable business to continue. This feature is often referred to as network continuity. This kind of failover scenario is ideal for small businesses to keep going during an outage. It used to be a slow and suboptimal affair, but has become much better in recent years. Today, it is a good option for smaller organizations.
For larger businesses, however, network continuity might not provide the level of connectivity required. At a larger scale, with multiple carriers, tools start to get expensive. Unfortunately, lots of connectivity flows through the same hole in the ground because it is much cheaper to share one conduit than dig a second.
In fact, there doesn't even need to be a cable break or outage. Network circuits need downtime for maintenance, and that downtime isn't often included in a service-level agreement (SLA).
Disaster recovery is a key part of continuity
Cloud providers, large data centers and colocation facilities are acutely aware of these network continuity issues and their effects on customers and business. To mitigate against complete connectivity loss, most will offer a blend of bandwidth taken from several Tier I network providers. These providers have separate circuits so that if a carrier circuit goes down or is unavailable, traffic can still flow without major issues.
Cloud vendors such as Microsoft have many providers as well as internal diverse circuits to ensure that if they have to perform a recovery, they have that capability in excess.
As you move down the provider tiers, the consumer owns more of the disaster recovery responsibilities. Most DR vendors have realized that a lost network circuit by itself can be a reason to invoke a DR plan, so most good vendors have included the capability to completely run the process from either side of the network.
Invoking from the DR side is not a problem because all but the last few minutes of data after the disruptive event struck will be accessible. That is why most DR products use asynchronous near-time replication: It makes the data available with less data loss. If one side is lost, the other can still work.
Cable breaks happen, and how long they take to fix can vary wildly. This time frame should be detailed in an organization's SLA. If a company does need to enact a DR plan, there are no half measures. The recovery software should be able to bring up those key applications simply and efficiently. All the steps and processes should be noted in the runbook for each application.