Protecting the enterprise remains a big challenge facing data centers. Backup touches and interacts with almost every component of the data center, and in the enterprise where there are more of those components; the chance for failure is high. At the same time, the expectation for rapid and frequent protection is increasing, as is the expectation of a rapid recovery.
Organizations are looking to the cloud as their means of escape, but for many enterprises, the cloud creates more problems than it solves. IT planners need to understand the potential problems when considering cloud backup as well as how to work around those problems.
Understanding the Cloud Options
The first step when deciding on a cloud backup strategy is to understand what the various cloud options are. The first set of options to understand is the software and hardware that will transfer data to the cloud. The second is to understand what type of cloud will the organization send its data to, and the third is what, if anything will/can be done with data residing in that cloud.
Almost every enterprise-class cloud solution will have some sort of on-premises requirement. That may be an appliance that sits on-site and which receives the backup before transferring it to the cloud. The on-premises appliance will more than likely require enough capacity to store all of the backups. The cloud is essentially a mirror image of the appliance. Other solutions can offload older backups to only be in the cloud, reducing the on-premises data protection storage investment. The on-premises requirement might also be installing software either on a dedicated server or onto existing servers.
Some of the appliances and software come as turnkey solutions, completely replacing the existing backup infrastructure. Others are add-ons to the existing software and hardware.
The second consideration, what type of cloud should the enterprise use is also critical. Providers range from the “megacloud” providers like Amazon, Azure, and Google, to relatively small regional providers. The megacloud provider may not be able to provide the enterprise with the customized service it is expecting or requires. The regional providers may be too small, lacking sufficient compute and storage capacity to provide adequate enterprise-class services. They may also lack a secondary site to replicate data to so they can meet the organization’s disaster recovery expectations.
In between megacloud providers and regional cloud providers are custom providers that have a global footprint and offer a seamless cloud experience but still deliver localized support allowing customers a choice as to what software applications they can use on-premises and in the cloud.
They don’t force the customer to switch to a “cloud-native” application. These custom cloud providers can also work directly with customers to meet the organization’s service level agreements (SLAs) around compliance, data governance, and security. They can provide custom “as-a-service” functionality based on the existing software and SLA expectations, something that the megacloud providers and most regional providers can’t or won’t.
Finally, IT needs to understand what it can do with its data once it lands at the cloud provider. In many cases that data is stored in the backup format of the protecting application. Action on that data, whether in the cloud or on-premises, requires recovery by the backup application into either the cloud compute layer or on-premises compute. Other applications store data in a native format that is immediately accessible by cloud or on-premises compute.
The 3-2-1 Rule Still Applies in the Cloud
The 3-2-1 rule states that an organization should maintain data availability by having at least three copies of all data, on two different types of media and at least one of the three copies should be off-site. This rule which has been around for decades still applies to the cloud era. The cloud is an ideal location for that one off-site copy, and with replication, it is easy to have two off-site copies in two different locations. Considering that cloud storage and on-premises storage both use hard disk technology at their core, they are very different in how they use those hard disks. Most organizations can count on the cloud as one of the two different types of media.
Also, organizations need to apply the 3-2-1 principle to cloud-based applications and data. Software as Service (SaaS) solutions like Office 365 and Salesforce.com are increasing in popularity, but the same 3-2-1 rule should apply to their protection. The backups that SaaS vendors perform are only to protect themselves not their customers. For example, Microsoft clearly states in their marketing for Office 365 that it is “your data” and that “you own it”. Microsoft backs up 365 for protection from a disaster but the users of 365 are responsible for their own data protection. Organizations should make sure that a copy is backed up on-premises and to an alternate cloud provider.
For enterprises to consider cloud storage for backup they need more than a great price. Cloud providers need to offer enterprises the right level of service so that the use of cloud storage is seamless. They also need to respect basic principles like the 3-2-1 rule of data protection. In our next column we will discuss how enterprises can select the right cloud for them.