Data Protection Workflow

I was speaking with Bocada and Tek-Tools last week about their focus onhelping customers solve their data protection challenges. If you thinkabout it there is no shortage of solutions to protect your environment,and I am encountering more than a few data centers that have the samedata set protected four or five times. Ironically, they don't feelanymore comfortable about recovery. The problem is a lack ofprocess or workflow to all these point solutions.

This lack of confidence leads to multiple data protection methods such as snapshots, continuous data protection, applicationspecific backup, environment specific backup, enterprise backup andreplication, to name a few. The lack of a workflow or process leads to aduplication (or triplication) of effort. It also leads to not knowingwhat protected copy to turn to first in the case of a failure. The lackof a workflow may also lead to a resistance to continuous testingbecause without a workflow, knowing what part of recovery to test can bedaunting.

The large vendor response to this has been to try to consolidate asmuch of this data protection as possible into a single package;Enterprise Data Protection, and there is nothing wrong with that ifthat vendor can cover all your specific needs. Despite the enterpriseprotection vendor's best efforts many environments opt for additionalprotection. Look at the rise in application specific protection toolslike those from Kroll Ontrack and AppAssure or platform specific toolslike those from Vizioncore or Veeam. Finally, there often is anoperating system itself that is awarded special treatment and theadministrator is allowed to use their own tool.

The reality is that no matter how good and how complete the enterprise dataprotection applications become, you will have other data protectionprocesses in your environment. What's needed is a unifying tool thatcan help you build your own workflow around data protection. There isnothing wrong with protecting something more than once as long as youknow its happening, and as long you know which protection copy you aregoing to count on for recovery.

The workflow should of course inventory all the data protection stepsthat are occurring in the environment, then it should help you identifywhat can be eliminated. For example do you really need Exchange backedup twice? If you are using an Exchange specific tool, can you have yourbackup application back up the data that the exchange specific toolcreates? There are pros and cons to each choice. Developing a workflow allowsyou to decide which is best.With a workflow defined, you now have something to monitor to make sureit is all working. This could be automated through a monitoring toolthat could support various data protection processes, but even manualauditing of the process would be greatly sped up by following aworkflow.

Finally, part of the workflow would be what to do in a recovery. Itshould be situationally based on the recovery conditions. Is this a test,a disaster or something in between? All the circumstances that couldcause a failure should be defined and a workflow developed.

Developing a data protection workflow will improve the quality of theprotected data, could reduce costs associated with data protection andshould reduce the amount of time spent dealing with data protection.All of this should lead to improved and more frequent testing which leads to improved confidence to recover from any type ofdisaster.