If you’re like most companies today, your disaster recovery (DR) plans rely on disk-to-disk (D2D) backup to provide recovery of your most critical data, applications, and systems. While D2D provides fast, on-premises recovery speed and availability, it also has some drawbacks.
First off, space is always an issue with traditional D2D SAN/NAS appliances. In order to meet DR objectives you need to correctly size the local appliance, typically 2x the amount of data. If the goal is to protect 10 TBs, the appliance needs to be 20 TB. And if a second copy is desired, a second backup appliance would need to be purchased, making the appliance storage required equal to 4x the amount of data – and 4x the cost.
Additionally, D2D’s real DR value is limited to disasters where the original physical location is still intact. For disasters that make your business location unusable, you either need to take your SAN or NAS with you, or have an alternate DR strategy, such as a hot or cold site lying in wait.
These drawbacks aren’t new and have some of your peers looking at the cloud for better methods of DR. By taking a hybrid approach to cloud backup, you can enjoy the improved availability, lowered costs, and ease of use the cloud brings, while still taking advantage of on-premises D2D backups.
If you’re considering moving to the cloud, there are several emerging trends that may push you over the edge by making the cloud a more viable option to on-premises D2D.
So, what’s new that’s improving the cloud (and your ability to utilize it) for DR?
Hybrid Appliances Are Paving the Way
You’re comfortable with the concept of D2D backups, so to simplify the transition to cloud-based DR, new hybrid (or “cloud-connected”) appliances are emerging that act as your local D2D backup, and facilitate the creation of a 2nd copy of your data to reside in the cloud. In addition to creating data redundancy, using these hybrid-cloud backup appliances also improves recoverability with or without the availability of the local appliance.
Despite the improvement, using a hybrid appliance still brings challenges of its own. Storage in the cloud is theoretically infinite (or at least should be, depending on the cloud backup vendor you utilize), but local storage, by definition, is limited to the size and number of local disks. Since hybrid appliances still use traditional appliance-to-appliance replication, they enforce a 1:1 replication of backups from the cloud-connected appliance to the cloud, and thereby artificially cap what you can put into the cloud. Some companies are changing the 1:1 paradigm by decoupling cloud storage from appliance-based storage.
Intelligent Caching is Improving Recoverability
Even with the introduction of hybrid appliances, the focus is still on the local appliance and its capacity limit sets the tone on how useful the cloud will be. What’s happening next is a shift that makes the cloud the DR focal point, where the local appliance acts as an extension of the cloud.
The next generation of appliances act more like a cache than storage, they keep everything in the cloud and also keep your most critical data (for example - the most recent backup, accounting files, the CRM database, etc.) local.
Using this type of appliance yields two huge advantages: Fast, local recovery of what’s truly important is available at a moment’s notice without the need to obtain data from cloud backups. Marry that with a removal of the 1:1 storage limitation of regular hybrid appliances, and you now have a fast recovery solution that allows the cloud to be a source of infinite storage.
With this unlimited ability to store all your data in the cloud, you can see the potential for putting too much in the cloud. Even with the most critical data sets stored locally, some may be concerned that you can’t easily get backups in the cloud recovered as rapidly as their current D2D solution.
So, how can you have confidence that everything you put in the cloud can quickly be retrieved during recovery?
Better WAN Optimization is Speeding Up the Cloud
If you’re going to stream more data to and, more importantly, extract data back from the cloud, the process of moving data to/from the cloud has to be fast. Historically, transporting data to/from the cloud is slow and completely dependent on the speed of your Internet connection. You can’t very well implement cloud-based backup and then increase your Internet pipe every time you want to backup some additional data set. That’s where new advancements in WAN optimization come into play.
It used to just be simple compression technologies that allowed more data to go through the same pipe. But today’s optimization efforts go much further with improvements such as traffic shaping (where you can define when back-up data is allowed to go across the wire), block-level in-line deduplication (to minimize the amount of changed data that is sent across the wire), data compression, and resilient-resumption of backup and recovery operations in the event of network disruption.
Couple WAN optimization with the first two trends and you have a “best of both worlds” (local and cloud) recovery solution where both the data stored locally and the data transmitted to the cloud are all optimized, making fast recovery of everything you need from the cloud a reality.
But what happens when a disaster strikes, either man-made or some type of natural disaster? How do you quickly recover mission-critical data and applications with minimal business interruption?
Disaster Recovery as a Service
In the event of a disaster that takes your building with it (and your local D2D backups, as previously mentioned), the only recourse is going to involve rebuilding your critical data, applications and servers at a remote location. For many companies with only an on-premise backup appliance, they’ve lost their backups entirely and it’s game over. But, with hybrid appliances the possibility opens up for disaster recovery as a service (or DRaaS for short).
Fundamentally, DRaaS is about allowing IT admins to bring back lost infrastructure in minutes without the need to setup new servers, download data from the cloud, or even re-install applications. If the environment was already running in a hypervisor, then that’s great. If not, then the environment is virtualized as part of the backup procedure, literally a physical to virtual conversion. In both cases, the environment is booted up either right on the backup appliance (assuming it was offsite) or directly in the cloud. Users can immediately resume operational use. When IT rebuilds the production infrastructure, the virtual environments can then be moved back to the corporate network seamlessly.
All of this sounds great, but there’s always one last question every customer is going to ask: “What is it going to cost me?”
Cost Efficiency is Driving Adoption
When you try to pitch your boss the promise of fast, unlimited backups in the cloud, and an optimized local appliance to make recovery more reliable and less time consuming, your boss may just be hearing dollar signs cha-chinging in his or her ears.
In reality, using a cloud-connected appliance with intelligent caching, along with cloud storage is far less expensive than continually trying to keep up with the 2x storage paradigm. And way less expensive than buying another, bigger, beefier back-up appliance. Buying more cloud storage is vastly less expensive and far more practical. It also shifts company spend from large one-time capital purchases to much lower monthly operating expenses.
It’s Time to Shift the Center of your DR
The cloud is no longer this mysterious undefined concept full of hype and promise; it’s a cost-effective means to solve real business problems, including DR. Back-up experts are embracing the cloud, developing ways to provide you with cloud-based DR with on-premises performance and availability. It’s time you embrace DR in the cloud as well.
Ken Shaw Jr. serves as CEO and CTO of Infrascale. Shaw is an experienced engineer and entrepreneur in the cloud storage and online backup industry. His expertise in storage oriented cloud computing and offshore software development has led him to be a thought leader in the cloud storage industry.