Your IT enterprise perimeter and surface exposure has become much greater — there are employees, customers and partners, cloud and on-prem, IoT and apps, systems and data, mobile devices and remote access. It's estimated that more than seven billion people and businesses and at least 30 billion devices are connected to the Internet, and data is growing too — it's expected to hit 100 billion terabytes by 2025.
With all that digital activity and noise, building or rebuilding your Disaster Recovery (DR) plan is critical. The place to start is by determining your Recovery Point Objective (RPO) and Recovery Time Objective (RTO). These key parameters will help you understand your downtime costs compared to your DR costs. Finding the right balance between RPO and RTO is critical to determining the technology that meets your needs and your DR budget, whether it's backup, replication or a mix of both.
Backup and replication — what’s the best approach?
Backup involves making a copy or copies of data. It's a relatively inexpensive way to avoid complete data loss. It relies on snapshots of the data taken at predetermined points in time. It requires a long-term storage location in a secure place. Backup typically is used to duplicate everything in the enterprise and for long-term archival of business records. While valuable for historical purposes and compliance, it doesn't ensure the smooth continuity of operations.
Replication, however, can ensure that data and processes are always available, even after an outage. It is the act of copying and then moving data between a company’s data sites. It requires investment in a second infrastructure, so it's more expensive, but it truly is disaster recovery because it enables quick and easy resumption of operations. There's a VM that's continuously replicating and just waiting to be switched over if needed.
RPO — the amount of data to be recovered.
RPO refers to how much data is at risk using a particular method. Here's an example. Let's say a company runs a full backup of their data every hour, and their data gets changed regularly throughout those 60 minutes. If they run their backup at noon on the button and suddenly their systems fail at 12:01, they've only lost a minute of live data. But if everything is running smoothly until a failure at 12:59, they've lost 59 minutes of data changes since the last backup. An organization with data that changes very slowly may be fine with a 24-hour RPO if changes lost in the last 24 hours won’t have a major impact. For businesses with financial transaction or medical data, even 10 minutes` worth of data loss is too damaging, so the RPO has to be much shorter because the risk is too great.
RTO — the amount of time to recover the data.
RTO is all about how long it's going to take to get lost data back into a consumable format so you can be up and running again. RTO helps you determine the method and technologies used. If you go with backups, how many backups is it going to take, and how long?
A backup takes all of the data, de-duplicates it and compresses it all together into a file, and that file goes onto another storage device. To recover the backup, you have to retrieve the data, rehydrate it, and recover it to the replacement server (physical or virtual) to make it usable again.
Compared to backup, replication is lightning fast in terms of RTO. It takes a very short amount of time to recover a failed virtual server, instead of having to wait for the backup to be recovered using the traditional backup methods. With replication, the changes made to a live virtual server are copied to a secondary location. In the event of a failure, the secondary site can be brought up (failed over) and the server can continue its functions much more quickly.
So, as you look at your DR plan, RTO is critical to consider — what data, systems and applications need to be available within minutes, hours or days? If you need all your data instantly because it's mission critical, you'll want 100% replication because it's real time or as close to real time as possible. You flip the switch, turn on the VM, and you're back in business. However, if you have different levels of data and applications needs, you can opt for a plan that includes both replication and backup.
The DR solution that meets your needs and budget.
OVHcloud® Hosted Private Cloud gives you maximum performance and optimum security at the right price. You’ll know exactly how much you're going to spend on storage and per VM. There are no commitments or hidden charges for network usage and no ingress, egress or API call fees. There are no zone, region or bandwidth charges.
To learn more about the competitive advantages OVHcloud Hosted Private Cloud and best-in-market RPO and RTO solutions, visit us.ovhcloud.com.
1 - Seagate Study: "The Digitization of the World from Edge to Core"