The success of cloud providers like Amazon, Azure and Google is forcing the data center to re-think various aspects of their IT infrastructure. There are cases where a cloud provider can enhance typical on-premise storage to be DR Ready and make long term retention viable.
Most organizations hope to never or at least rarely need archives; however, because of the implications to the business of not preserving their data, it is something they have to invest in. That investment may include development of another facility, along with the effort of purchasing or renting the physical location and then putting in all of the equipment.
The cloud has the potential to lower the “rent” and eliminate acquisition costs. However, there are cautions.
Data Has Gravity
It might be easy and even free to import data into public cloud but could cost a great deal to try to get any data back. On top of that, you are compromising on the level of control by choosing to pay the “rent”. IT professionals need to weigh the costs of quick startup vs. the long term costs of renting storage. IT also has to be careful before jumping on the cloud DR bandwagon, the cloud’s DR site capabilities are not yet up to 99.9999% SLA required by many enterprises. The key is flexibility, a DR Ready storage system should be cloud-aware so that the organization can choose to use the cloud in a way that makes the most sense for it.
While many on-premises storage systems claim to have the ability to integrate with a cloud provider, it is a kludgy afterthought, forcing IT professionals that want to leverage cloud economics to find a third party solution, even though their existing storage system is oh-so-close to being cloud ready. It’s wise to take a closer look how that connectivity can lead to the savings you originally intended to.
What Can You Do With Your Cloud Connection
Every primary storage system worth its salt has snapshot technology and most of those leverage that snapshot technology to replicate data to a disaster recovery site. The problem is the target system almost always has to be a near identical system. It is, of course, unlikely cloud providers will implement one of every storage system into their infrastructure.
What if the primary storage solution has the ability to replicate snapshots to an simple storage service (S3) target. S3 is the API Amazon provides to interface with its object storage, and it has become the defacto standard for object storage interfaces. Almost every cloud provider and most object storage systems are either native S3 or provide an S3 interface.
With an S3 interface some or all of an organization’s data could replicate to a cloud provider instead of, or in, addition to a secondary system in another facility.
The Value of Cloud Aware
Once this data is in the cloud, an organization could likely “check-off” their long term retention requirement. However, from a disaster recovery perspective, IT professionals should be more careful. Replicating to a secondary site with a similar system as opposed to the cloud may still be the preferred option. A secondary site is under their control and has a known enterprise class storage system. A quality system at one or multiple alternate sites is critical as those systems may one day become the primary in the event of a disaster. In this situation the cloud is still attractive as a location to get data “when all else fails”.
For organizations where a second site is not available then the cloud as a DR site remains an option. Assuming they work out the obvious network configuration changes, recovery from a disaster could occur quickly without having to wait for the implementation of new equipment or storage. They will have to transform their VMs from the hypervisor that they use on-premises to the hypervisor that the cloud provider uses. While the organization saves the upfront costs of having to physically buy facilities and physical equipment there is more work required to ensure that the failover process will work and is understood.
While cloud integration is important, “lift and shift” your entire DR to cloud is another extreme. Organizations with secondary sites available to them may want to look at those sites as the first option and the cloud as the recovery point of last resort.
Long term data retention has become one of the best use cases of the cloud. The economic and logistical advantages are very compelling. Using the cloud for DR is another option especially for single site organizations. Organizations with multiple sites should look for storage systems that provide the ability to replicate to those sites while also being cloud aware. Then these organizations can use the cloud for long term data retention and as a backup to the DR site. In the end it is all about flexibility. Your primary storage should provide the flexibility to leverage and integrate with the cloud for both DR and long term retention models.
Sponsored by Tintri