Creating a Server Disaster Recovery Plan

We all know that the world isn’t perfect, and disaster can strike at any moment. This can be a natural disaster, a power outage, a fire, a flood, physical fault in the server caused by one or two hard disk failures, a ransomware attack—disasters take many forms.

Sometimes your small business clients might need to be reminded of that. If it catches your client unaware, a server crash can not only grind their business to a halt, but also lead to the loss of valuable (and irreplaceable) data. However, with a server disaster recovery plan in place you can help mitigate that downtime and prevent any data loss.

Every business and organization, regardless of size, needs a server disaster recovery plan. A good server disaster recovery plan saves your clients from the pain that comes from having all of your eggs in one basket—like a trip to our data recovery lab so that our RAID server data recovery experts can get you back on your feet.

Backup: The Heart of Server Disaster Recovery

At the heart of any good server disaster recovery plan is backup. Whether your server lives in a cloud or a closet, you need to make sure everything is redundant.

Of course, your backup system won’t always be an entirely separate physical server—especially not in this day and age. Up until relatively recently, server disaster recovery would often involve bringing out a backup of the affected server—on tape. This type of backup, of course, took time to acquire and integrate. Nowadays, it’s more common to use server virtualization as a form of disaster recovery. Physical backups are unwieldy and expensive. Virtual and cloud-based backups are much cheaper, more convenient, and easier to maintain. Not to mention, they’re easier and faster to transition over to as well.

For example, with a full-image backup service, your small business client can transition over to a virtual server right after your server takes a dive with very little appreciable downtime. While you run off of the backup server, you can set to work getting your physical server back up and running.

The Dangers of the Cloud

Many small businesses try to conserve money and resources by leaving everything up to the Cloud. With so many reliable cloud-based hosting services out there, it seems like a safe bet. Far too many people say, “These reputable cloud hosting companies all have their stuff together, don’t they? I don’t need to back up my stuff somewhere else if it lives in the cloud.”

Of course, nobody’s perfect. Even Amazon S3 had an uncharacteristic outage that affected thousands of websites for five hours not too long ago. A lot of people found themselves up the creek when this happened because they thought using Amazon Web Services was a server disaster recovery plan in itself. There were certainly some businesses out there, though, who, thanks to the advice of their IT consultants and managed service providers, had a backup plan when Amazon’s service faltered. (And those IT consultants and MSPs certainly deserve plenty of praise for that.)

Of course, there’s absolutely nothing wrong with using cloud-based services. Even some of Gillware’s own servers live on the Cloud! In fact, cloud-based backup is an essential part of a modern-day server disaster recovery plan. But your clients need to know that they can’t count on their cloud-based server being any more failure-proof than the server rack standing in their closet. Don’t let the Cloud lull your clients into a false sense of security. The Cloud is a powerful tool for backup. That said, it does not absolve you or your clients of responsibility when it comes to backups.

What a Server Disaster Recovery Plan Looks Like

There are a few things to keep in mind when coming up with a server disaster recovery plan for your small business clients. Mainly, you want the steps that need to be taken documented well enough that even a junior IT technician they hired that very morning can follow them. After all, disasters can strike at any time, and your senior IT technician might be on vacation in the Bahamas or in the hospital with a broken leg on that day. You just never know.

Our hypothetical IT newbie needs to be able to skim over the plans and immediately know where the server closet is and how to get to it. When Junior gets into the server closet, they should be able to figure out relatively easily which server does what. They shouldn’t need to ask a coworker who’s been there longer than them which server does what and hear “Well, Ol’ Silver manages our phone system, Big Blue is the email server, and Deep Thought has our CRM,” only to get into the server room and have to spend valuable time figuring out which of your racks is Ol’ Silver, which is Big Blue, and which is Deep Thought. (We’re operating under the assumption that, of course, all the servers have kitschy names.)

If getting to the server is exceptionally difficult, then your server disaster recovery plan should at least provide a decent walk-through. Once Junior has rotated the bust of Aristotle 270 degrees counterclockwise, solved the revolving door puzzle, and gotten into the server room, your plan needs to clearly detail the steps they need to take to smoothly transition over to your backup systems.

Questions Your Plan Needs to Answer

For every server your client uses, you should have a plan for its eventual (and inevitable) failure. For example, Gillware has over a dozen servers. Some of them live in our closet. Some of them live on the cloud. All of them have an entry in our disaster recovery plan.

Here are the basic question your server disaster recovery plan needs to answer for each of your client’s servers:

Where do your servers live?

• What does the server do?
• Where does the server live?
• If it’s a physical server in your building, what does it look like, and what is its brand and serial number?

Who knows how to get to your servers?

• Which employees have the keys to your server room and/or the access credentials to your cloud-based storage accounts? Don’t keep your login credentials in your plan itself—just record who in your organization knows the credentials.

Where do your backups live?

• Does its backup live on the Cloud, or in a failover box elsewhere in your building?
• How do you tell the backup to take over?

Planning to Avoid Disaster

For a small business owner, a good server disaster recovery plan can get you back up and running in no less than just a few hours, if not near-instantaneously. For an MSP, when your disaster recovery plan pays off, it has the added benefit of reminding your client just how valuable your IT consulting services are. When you keep disaster recovery in mind when setting up your servers and your backups, your disaster recovery plan can also make it less likely for a disaster to strike.

Keep these things in mind when deciding how to set up your server and its backups:

Where do your backups live, and who can access them?

To keep your backups safe and secure, keep them off the network. This is especially pertinent when planning to protect your data from man-made disasters such as cyberattacks and ransomware intrusions, in which a bad actor will attempt to get onto your network and make mischief (with serious consequences for you). If your backups live right next to your server on the same network, an intruder can really ruin your day.

Ideally, only one trusted person, connected with your backup service provider if you use a provider such as Gillware Data Solutions, should be able to access your backups and make changes. This is one of the advantages of cloud-based, offsite server backup services. Any local backups you do keep should be protected by a firewall.

How often do you update your backups?

Your backup process should be automated and incremental. You should have a “living” backup process that regularly scans your servers and updates with every change that gets made. This way your backups won’t become out-of-date and suffer from “configuration drift” as your actual server changes.

In Gillware’s data recovery lab, we’ve seen RAID servers come to us from small businesses that had backups, but when disaster struck, the backups they had of their server no longer accurately reflected the settings and configurations of the server, making the backups all but useless.

How do you report on the health of your server?

You should have some way to monitor the “heartbeat” of each of your servers. If a server starts flatlining, you want to know right away so you can implement your server disaster recovery plan posthaste. For this, you’ll want to go with an automated service like Nagios that can keep a close eye 24/7/365.

When setting up your server, you should also look into how it can notify you of its status. Having IT personnel regularly check up on your server to make sure all of its disks are healthy and replace any that need replacing is infinitely preferable to leaving your server in a closet to gather dust.

There are other ways to monitor your server’s health. Some consumer-grade NAS devices frequently employed by small business owners and freelancers as cost-effective servers can automatically notify the owner when one hard drive fails and needs replacing, for example. However, these devices rarely come configured “out of the box” to do so. As a result, these NAS servers can suffer preventable crashes, seemingly without warning.

Creating a Recovery Plan

One of the most important things any of your clients can do is create a disaster recovery plan for their business. A good disaster recovery plan can minimize both downtime and data loss in the event of anything from a server crash to a ransomware virus or other form of cyberattack.