One of the reasons why disaster recovery attempts often fail is DR process planning and operation is so expensive. The expenses associated with the DR process leads organizations to try to cut corners in an effort to reduce costs. That corner cutting often leads to holes in the disaster recovery plan that then leads to a failed recovery or a recovery that exceeds organizational objectives. Organizations need to re-think their disaster recovery process and intelligently leverage the cloud so they can cut recovery expenses while speeding the recovery effort.
Building the Perfect DR Site
Several years ago we were involved in a project where the disaster recover site was a perfect match to the primary data center. Everything purchased in the primary site was also purchased for the DR site. When something was removed from the primary site it was removed from the DR site. All primary data was replicated between the two sites via a high-speed connection, the DR site was never more than a few minutes out of sync with the primary site. Even though there was no disaster, twice a month we failed from the primary site to the secondary site. The secondary site would run operations for a few days, we’d reverse the replication jobs, and then fail-back to the primary site.
Then a real disaster hit. The downtown area, where the primary site was located, was flooded, within seconds we had operations moved to the DR site and ran operations out of that site for a few months. When it was time to move back to the primary site we were able to do so seamlessly. It was about as perfect a DR site as you can get…it was also incredibly expensive.
The reality is that most organizations are not going to be able to afford the “perfect” DR site. But organizations that leverage the cloud can get very close to perfection and for a lot less money than they might think.
DRaaS is a Good Start
Disaster Recovery as a Service (DRaaS) solutions typically either backup or replicate an organization’s data to a cloud provider and then in the event of a disaster, enable that organization to run its applications as virtual machines within the provider’s cloud. DRaaS eliminates some of the biggest expenses related to disaster recovery. First, there is no need to own the DR site. Second, there is no need to own the equipment (servers, networking) at the DR site. Third, a continuous testing model like we described in the Perfect DR site is very easy to do. These capabilities have made DRaaS one of the fastest growing service offerings on the market today. Every organization, regardless of size should consider DRaaS as a key part of their DR process.
Where DRaaS Falls Short
Missing from the list of areas where DRaaS saves money is storage. While it does help, at least initially, to be able to “rent” storage, the total cost of capacity does start to add up. Most DRaaS solutions have an on-premises component to them to store local backups and most organizations use some other process to handle copy data – data that is used to feed DevOps, analytics and reporting processes. The result is that in addition to the “rented” storage organizations still have a large and increasing on-premises storage bill.
There is also a return element to DRaaS that is not as smooth as it could be. In most cases the organization, after the disaster, will either want to, or have to leave the provider’s facility and start operations back in their primary facility. This means a big lift and shift of data, which can be time consuming and error prone.
The Cloud as Primary
The next step in the development of DRaaS solutions may be to use the cloud as the only data store. Essentially, the organization stores all of its data in the cloud and pushes it to the primary data center on an as needed basis. A high performance edge device acts as a cache for that pushed data so that when that it is in use, performance is up to the organization’s standards.
The challenge with the cloud as primary storage is, what happens when their is a cache miss on the edge device? The latency of a cloud fetch will be noticeable. The solution is an intermediate storage area, a metro point of presence that can respond to a cache miss in milliseconds instead of seconds.
With this type of solution in place, the cloud can be leveraged for copy data, and there is really no need to “return”, simply reinstall the edge device and the organization is up and running.
To learn more about leveraging DRaaS and Cloud Primary Storage to reduce disaster recovery costs, watch our on demand webinar.