Backup has been constant throughout my career. From the days of big open reel tape drives, massive tape libraries
to the Virtual Tape Library (VTL). Having survived several IT disasters, most of which were man-made, I very quickly
learned that backups – need to complete and be complete,but your credibility as a data centre manager is measured
when its time to restore, or recover.
Time and Point are the important objectives. Meaning how long will it take to recover and at what point in time can I
get back to ? Answers to these questions became a key element of corporate data governance and protection. As
workloads generating new business value replaced older workloads the analysis of both backup and recovery objectives
became the cornerstone of the collaboration between IT and the business.
Over the last 30 Years backup and recovery has evolved. Back then it was a hugely labour intensive activity. It took
up acres of floorspace and was, without a doubt, error-prone. Automation changed all that and one large organisation,
no longer with us, used robotics to change the dynamics within the datacentre. No longer were teams of shift workers
needed to manage tape mounts. The robot integrated with a large tape library automatically and
mounted the right cartridge on the right drive in good time.
But while robotic automation changed the dynamics and cost model within the datacentre. The media management
remained labour intensive. Remote vaulting of media was still required. We often discussed the cost of a Man-in-a-Van,
as a method of remote vaulting, versus network replication of key data assets. The ‘Man-in-a-Van’ option won as the
network capacity, cost and speed was often no match for physical transport.
Backup applications advanced and combined back & restore with library, media and data management. Useful features
arrived that backed up only the changes, differences or deltas since the last backup. This drove new efficiencies and
then came the slow demise of tape as the primary backup target. Along came disk based storage at the right cost often
emulating the tape library.
Whatever your method of backup and restore just remember one truth, and that’s Garbage-In-Garbage-Out, or GIGO.
No matter how good your backups are data corruption is always there to keep you grounded and displace any complacency.
This was a particular favourite in databases where a corrupt transaction would cause the application to baulk. The ability to
quickly restore from a recent SnapShot of the database and replay the transactions from the redo logs
(if you are running in that mode) kept the business running.
The process of backup has continued to evolve taking what is often considered to be a costly, mundane, but essential data
protection insurance policy and driving more efficiency and protection capabilities into the process. Now it is a core function
of any Data Governanace and Stewardship policy. Efficiencies and protection such as zero capacity copies, dedupe compression
and encryption are the expected norm and provide flexibility and choice.
Since the beginning of NetApp time, the early developers of ONTAP focused their minds on incorporating robust, rich
datamangement capability into the product. Today the ability to Cluster, SnapShot, Clone, Mirror and Vault data assets are
core functions within ONTAP. This capability was quickly incorporated into a host of Snap Managers for critical business
applications such as Oracle, MS Exchange, MS SQL Server, SAP and Hyper-V. There are many more, but to give you an
example, SnapManager for Oracle Integrates with, and automates, complex and time consuming database backup, restore
and cloning activities. Just what is required when you are managing a GIGO event in the middle of the night.
The latest addition to the NetApp Data Protection Armoury is SteelStore. As covered in my last Blog SteelStore meets the
growing demand for secure, efficient backup in the cloud. Yes there are questions of risk regarding backing up to the cloud
so please take a look at the NetApp customers who are doing it today. Wright Pierce is a great example of an engineering
firm who have seen massive improvements over tape based backup such as higher levels of data security, faster restore
times and lower costs.
Finally please remember – You are only as good as your last backup. If you are looking to reduce business risks, costs and
improve efficiency then NetApp have ExaBytes of data under the protection of ONTAP and Steelstore today. Experience counts !