VMware specific backup solutions have been around for years, and most of them have added some form of cloud connectivity to their feature set. But, most of these options simply use the cloud as a digital dumping ground and don’t fully exploit the capabilities that the cloud can provide. As a result, customers are wasting time and money and not getting the full benefit of the cloud.
Cloud Storage is Just the Beginning
The reality of the unrelenting growth of primary storage capacity has an exponential impact on backup storage since the secondary data set is typically five to ten times the amount of primary storage. That means even a 10TB expansion of primary storage can lead to a 50 to 100TB of secondary storage. That secondary storage, then, also needs to be replicated to an off-site location, which means another 50 to 100TB of growth at the secondary facility.
For most IT professionals the immediate appeal of the cloud is that it eases the pressure on secondary storage expansion. While secondary storage capacity still expands, the data center does not need to house all of it. At least that is the hope.
The reality is that most on-premises backup solutions, if they can connect to the cloud, can only make a cloud copy, and they can intelligently tier older backups to the cloud. They also can’t support the various tiers within the cloud like leveraging Amazon S3 for recent backups and Amazon Glacier for older ones. As a result, the client doesn’t reduce on-premises storage cost and can’t cost optimize the cloud-based copy.
Cloud for Disaster Recovery
There is also a lot of discussion about disaster recovery in the cloud. Known as disaster recovery as a service (DRaaS), the concept is to leverage cloud compute to instantiate virtual machines. Most VMware specific solutions don’t have to integrate the DRaaS capability they are trying to provide via service providers who manually have to transform and start up a client’s virtual machines in the event of a disaster.
Cloud for Backup Scale
Finally, most backup solution vendors don’t use the cloud to scale backup operations. Over time, as the amount of data under management grows, the databases that manage the backup data becomes quite large. These databases are used for a variety of tasks including how many versions of a data set are available, and where each of those versions are stored. They can also contain specific meta-data to make finding data years after it was first copied, much easier. As these databases grow, various operations applied against them begin to slow down. Eventually, the database becomes unusable, and the customer is forced to either prune the database down, which means keeping less history or they have to upgrade the backup server.
At the heart of the problem is that most backup solutions are not scale-out in nature. They can’t scale across nodes and leverage the compute of multiple nodes to be able to manage a robust and large database. This also means running the backup process in the cloud won’t help, since the cloud is designed from the outset to host scale-out applications.
True Cloud Backup for VMware
True cloud backup for VMware starts with a backup software solution that is scale-out and can take advantage of the cloud’s scale-out compute infrastructure. Doing so enables the product to retain virtually unlimited versions of data sets and keeps those versions as long as is requires.
The second step is to leverage cloud storage to store all the secondary data. Storing data in the cloud relieves the organization from having to build out a secondary storage infrastructure on-premises. A critical aspect, though, is for the true cloud solution to support multiple tiers of storage available from the cloud provider so that retained data can be stored at prices that are less expensive than even tape storage.
Finally, the true cloud backup solution should leverage cloud compute for recovery or DRaaS. In the event of a data center outage, the customer should be able to leverage cloud compute to start their virtual machines in the cloud, returning service to users and customers almost instantly. The benefit of instantiating a VM in the cloud though goes far beyond the DRaaS use case. The capability can be used to easily and regularly test the DR process, to leverage a cloud VM for test-dev and to leverage a cloud VM to run reports or perform analytics.
There are many vendors claiming support of the cloud as part of their VMware Backup product. But, there is a difference between “support” and fully utilizing the cloud resources. Cloud compute should be used for scaling backup operations as well as hosting customer VMs during a disaster. Cloud storage should be fully utilized, including the various tiers of storage provided by the cloud provider.
To sort through the cloud backup maze, register for our webinar “The Best VMware Cloud Backup and DR Options”. In this webinar, you’ll learn the five requirements for true cloud-based VMware backup & disaster recovery.