Updated for May 2017 (Originally posted December 2014): Do you run any part of your business in the cloud — maybe Office 365, Salesforce.com, ERP (Enterprise Resource Planning), payroll or voice calling? How about email? The more you have in the cloud, the more you need a working internet connection. You need your employees productive and your virtual doors open for business.
Businesses that use the cloud need redundant connectivity from diverse carriers on different mediums. Here’s what that means:
1. Redundant network connections — It’s not just your office that can go down. Your carrier can go down too.
Most carriers will experience an outage at least every 16 months. So you need to cover your own common points of failure, like firewalls and switches, plus have a backup plan in case your carrier goes down.
The internet isn’t a straight line — IP backbones are more like webs — and major points of presence (POPs) are regional. That’s why, for example, Hurricane Katrina affected Internet connections all over the country. Traffic that was routed through the Panhandle had to be redirected, which slowed down carrier customers that were nowhere near New Orleans.
The same concept applies when your main provider may not actually be down but is just slow. With redundant connections, your router or firewall can be configured to automatically and seamlessly switch from your primary connection to your backup connection until things get back to normal.
2. Diverse carriers — Because your main carrier can have a terrible, horrible, no good, very bad day.
Make sure not to use the same provider for your primary and redundant Internet service providers. Having both connections from Comcast, for example, isn’t true redundancy. What if the problem is the provider?
3. Different mediums — You need a viable detour on an entirely different road.
Depending on your area, your options for connectivity include:
- Fiber — Your best choice for your primary Internet service because it’s the fastest and has the newest networks with highest bandwidths. But it’s not available everywhere. You usually find fiber in major cities with dense office buildings.
- Cable — Your best plan for low cost and reliability and your best backup if you have access to fiber. Your business can probably get it for $200 to $400 a month (it would have cost you $1,000 a month 10 years ago).
- Copper lines — Super common way to access the Internet — includes T1, DSL, DS3 and others. However, the phone network was originally built for voice, not data, so converting signals creates inefficiencies and opportunities for failure.
- Wireless — If your business has the good fortune of being in the line of sight of a tall building that’s acting as a distribution point, or if you can use another wireless bridging technology such as satellite Internet, 3G/4G or WiMax, these are great options for backup. But they will probably be too slow to be a viable option for your primary connection.
Understanding these issues involved in connectivity is core to your cloud strategy. For example, is your business in an office building that’s wired with fiber? Then up to the cloud you may go. Or is your business in the boonies with phone lines as the only way to get online? Then in your data center you should stay.
Of course, it’s much more complicated than that, and most businesses fall somewhere in between. But regardless of how much — or how little — your business has in the cloud, you want N+1, which is geek-speak for having the number your business needs plus one redundant system. Think of your N+1 internet connection like a second car you keep in the driveway instead of trading in — just in case your primary car breaks down.
Leapfrog makes sure our clients’ internet is always available. That’s one of our primary roles in providing managed IT services. We help each client determine the best redundant connectivity/N+1 configuration for their needs, then do what it takes to get them there, including negotiating the best deals with their Internet Service Providers (ISPs). We have tools that monitor the network at all times and automatically switch traffic to the redundant system if the primary system hits a glitch. And if our clients’ ISPs aren’t delivering the availability and speed they promised, we step in to handle that, too.
If you liked this post, don’t forget to subscribe to FrogTalk, our monthly newsletter.