Cloud Data Backup: The low hanging fruit for cloud adoption

In simple terms, Cloud Data Backup is an approach by which enterprises can backup and restore data from a third-party cloud storage provider. For those of you familiar with data vaulting services from SunGard, IBM's BCRS, Iron Mountain, Cloud Data Backup may appear as a old idea being offered for more than a decade. So whats new? The biggest differences are the cost factor, the ease of setting up the service over the internet without upfront human consultation overheads, and the different operations/analytics you can do on your backup data while its idling away!  Most SMB and large customers already use data vaulting services -- these customers have little installed-base loyalty, and are looking for lower costs, ease of management, and better provenance of their data. I think this is truly a low hanging fruit towards cloud adoption.

Who needs it ?
I have come across several customer scenarios that may benefit -- some more relevant than others. Following a  few of my favorites:

• An SMB with a single data-center worried about regional disasters such as floods, fire, etc. Currently, data backup is done once a week on tapes, and shipped to a vault. To curtail costs, they comprise with RPO and RTO which are of the order of days.
• An enterprise will multiple small branch offices (ROBO). The updates are periodically synced with a central data-center using a combination of backup products.
• A use-case where the data needs to be restored to a site other than the original primary site.  
• An enterprise requiring for a tertiary backup solution or a backup in a different region/country.

Flavors of backup?
Cloud Backup offerings today come in different flavors.

• Periodic snapshot of data stored with the cloud provider. The snapshot frequency is typically a function of the RPO (Recovery Point Objective).
• Continuous replication of data from primary site. The updates are asynchronously committed to the cloud.

Also, the service provider may couple together storage and compute resources in the cloud, allowing various data operations on the backup data. A few examples of data operations include verifying the integrity of the data or running archiving policies on the data for compliance and e-discovery. Given the elasticity of the cloud, it is easy to imagine running a BI content mining application on the backup data.  


Challenges/Inhibitors?
Cloud computing is not a done deal. There are inhibitors that need to be addressed for broader adoption:
Security (in-flight and at-rest) and data governance, Ensuring data integrity and recover-ability of data, Ensuring performance guarantees especially critical for data restores (to ensure RTO goals), Seamless integration with existing backup software and appliance installations, Search and indexing of data across multiple versions. Also, lack of standards in this space cause fears of a single vendor lock-in. 

In my future posts, I will elaborate on some of the challenges and also opportunities for innovation for big vendors.