A secure network usually includes a number of redundant systems. In this video, you’ll learn more about redundant hardware and systems and how far the planning goes when designing redundancy.
<< Previous Video: Critical Systems and ComponentsNext: Continuity of Operations >>
Sometimes problem with business continuity occurs because you have a single point of failure. That obviously is something that if you lose a server, you lose a router, you lose something. That one thing, it can cause everything else to fail. That can really ruin your day, unless, of course, you’ve made plans for this.
Having that single point of failure can be mitigated. You might have additional hardware you put side by side. Maybe you’re making a network configuration of what we call the Noah’s ark of networking. You have two routers, you have two firewalls, you have two switches, and they’re all redundant. If you lose one, the network will still continue to function because you have a completely different piece of hardware right next to it that’s able to take over the load that’s going on.
And it’s not just networking you need to think about. You need to think about power. You need to think about your facility. You need to think about the cooling system in your data center. If you lost your cooling system, it will not take very long for the temperature to rise and for your computer systems to begin failing.
You also have to think about people and location, especially on things like disasters that deal with nature, hurricanes, for instance, in Florida, something everybody keeps in mind. And if a hurricane comes through, it could decimate an area. There could not be power for days or weeks. You might have buildings that are suddenly here one day and gone the next. How do you handle that?
Do you have people in a different location? Do you get a bus and you ship people somewhere to take over in a remote location for a temporary amount of time? It’s something you have to think about because that becomes a single point of failure for what you’re doing.
The reality is there’s no possible way you can remove every single single point of failure. There’s no way to do it. Money is really driving the redundancy. If you had all the money in the world, you could certainly create your own power plants, have completely separate power plants that are providing your particular building or multiple buildings with different power sources. Obviously, not everybody can build their own power plant. So at some point, your single points of failure can only be taken care of or mitigated in so many ways.
And if you keep throwing money at the problem, you can do a pretty good job of that. But ultimately, you have to think about and make a business decision about how far you can go with getting rid of every possible single point of failure. And somewhere in the middle, there’s a happy medium that everybody will agree on that will have redundancy in our network, will have redundancy for our facility. But at some point, we’re going to have to just rely on that single point coming in and maintain and try to make sure that we can work around that should a problem occur.
Here’s a good example of how a redundant network would work. For instance, you have multiple internet connections coming into multiple routers. Those routers are maybe using fault-tolerant firewalls, one that’s on standby and one that’s always used. And if one of those firewalls loses the magic smoke that’s inside of it, it comes out, and it doesn’t work any longer, you can fail over to the redundant system.
You might even have redundant core switches in your environment that are going to multiple servers. And even the servers themselves might have multiple network interface cards inside of them. So that’s a very good example, and it’s a very common example of how some of the biggest networks in the world are maintaining uptime and availability just by putting redundant systems in in the core and the edges of their networks.