Don’t fail from failovers: The difference between Failover and Disaster RecoveryDecember 14, 2018
Don’t fail from failovers: The difference between Failover and Disaster Recovery
It won’t only happen to others.
Imagine your IT systems go out.
It’s time to spring into action.
But wait – your backups are unavailable, locked in a storage room at another site, and you have employees who now aren’t able to work. Stress levels rise and profits drop.
No matter how you look at it, the quicker you can get your systems up & running again, the better.
Don’t let the word "disaster" lull you into a sense of security. A disaster isn’t just an ‘act of God’. While disaster can mean flooding and earthquakes and cyclones, more often it is something as simple as an unexpected update or patch for your system.
Recent research from disaster recovery provider Zerto, shows that more than a third (34%) of organisations across Asia-Pacific have experienced some sort of IT outage, disaster or major disruption. They also report that two-thirds of Asia-Pacific organisations have implemented a formal disaster recovery plan. Could you imagine being one of that third that has experience an outage, with no disaster recovery plan?
What is the difference between backup, disaster recovery, failover, and high availability?
Data backup is what it sounds like. You’re basically saving copies of your data at certain points in time, which could range from a few hours ago to a few days ago, depending on the specific organisation’s practice. This data is often written to physical tape, and although this is inexpensive, it is also unreliable (some numbers estimate that 50% of tapes fail to restore). It’s important to have a data backup solution in place, but it can’t be an isolated solution.
A failover is like a spare tire. When you get a flat tire, you switch to the spare and you keep driving. Some spares are only meant to take you to the nearest tire shop, but others are the same as the original tire. In the IT world, that generally means there’s a computer ready to run your mission critical programs and you can switch to that computer when your system goes down to continue working until the original system can be fixed by IT. Failover helps to preserve business continuity while IT fixes the problem.
High availability is a protocol designed for companies that just can’t operate with a disruption to their IT systems. Machines are configured to reduce any downtime and aims for little to no human interaction to restore operation to the system. High availability and failover are often confused because failover is typically how high availability is implemented, but they’re not the same and can be used independently.
Disaster avoidance can also be confused with high availability. High availability usually consists of a replicated version of the operating system in the same location (or data center), which means that if the data centre goes down, so do both working copies. With disaster avoidance solutions, the two mirrored systems are active but in different physical locations (usually a safe distance apart), which adds an extra layer of security and uptime guarantee. For instance, this is something Macquarie Cloud Services implements with their zero downtime hosting solution.
Disaster recovery is the key here. It covers all these ideas under one umbrella, that keeps systems working with little downtime while also providing a backup and recovery plan for when disaster strikes (operative word: recovery). A good disaster recovery plan will outline the procedures that happen when something takes down your system, whether it’s the whole thing or a small piece. It’s basically a handbook of steps to recover lost data during an outage, restoring it, and getting everything back on track. This might include replacement pieces of physical hardware, backup solutions (like tapes or backing up to the cloud), backup sites, data integrity checks, and more.
The only disaster recovery is… disaster recovery.
I don’t know how many times we’ve heard a client say, "Yeah, we’ve got a disaster recovery plan. We back up our data once a day." While that does save your data, it is not a solution for business continuity. Backing up your data does ensure that you can access it – eventually. But when you have a real disaster recovery plan in place, you can continue business as usual much, much quicker.
Think of it like this: you save all your personal data to the cloud – music, photos, anything you want to keep. Your laptop breaks so you purchase a new one. You can still get the data from your previous laptop, but only after setting up your new one, configuring it for your needs, and spending days downloading all that information from the cloud. Disaster recovery is like having a mirrored backup laptop that you can switch on, switch to, and be immediately back to where you were.
The three keys to any disaster recovery plan:
1. Easy to implement and follow. Follow the K-I-S-S analogy for this one (Keep, It, Simple, Stupid). It needs to be a plan that anyone can follow, just in case your expert is out of the office. A simple plan also reduces the chances of something else going wrong (the last thing you need).
2. Redundancy. Save your information in different ways and different places. For example, if something happens that affects the area where your office is, your data backed up on the cloud is still accessible, but physical copies might not be. Also, backups (even if they’re tested) can still fail. It is always good to have a backup of the backup.
3. Round-the-clock availability. Believe it or not, disasters aren’t polite enough to schedule themselves ahead of time. So make sure you can implement your plan at any time.
Thanks to the expert team at Macquarie Cloud Services for the contributed article around disaster recovery.