“every web service’s worst nightmare”

Red Door Interactive: Latest post

On Friday, January 30th, 2009, Ma.gnolia (http://ma.gnolia.com), a site that provides online bookmarking services, went down and the following excerpt was taken from a message posted on their home page:

“Ma.gnolia experienced every web service’s worst nightmare: data corruption and data loss.”

Let’s face it, you can own several Web sites for a lifetime and never experience a “nightmare”.  However, if it does happen how prepared are you to get operations back to normal? Backup (BU) and Disaster Recovery Plans (DRP) are often overlooked aspects of your Web strategy and I am surprised at the number of businesses that continue to run their operations without them.

Since disasters can range from someone deleting the wrong data to catastrophic events they are inevitable occurrences of our digital world.  Having experienced a few recovery situations, I have provided some key points and considerations for your BU and DRP plans.

Acquaint yourself with your data center or hosting facility.
Regardless if your site is hosted internally or externally, system administrators play a critical role in the backup and recovery processes.  You will want to meet with them to understand what your options are and get an understanding of your hosting environment.

  • Have you reviewed Service Level Agreements or policies to see if there any items that may affect your DRP?
  • What redundancies does the facility have for power and data storage?
  • Does the facility have a DRP?
  • Are there additional costs for including your hosting vendor in the DRP?

Understand your data.
Because the cost of BU/DRP is a function of how much data can be lost and how quickly it needs to be recovered, understanding your data is a major consideration when developing your plans.

  • Are there compliance, regulatory, or legal requirements for your data?
  • Does your data support e-Commerce transactions?
  • Should you consider high-availability recovery options to support real-time failover to a standby server?
  • What transactions rely on your data?

Test your plans.
If you do have BU/DRP plans, it is important that you verify the processes.  Your initial DRP may follow a “happy trail” of events.  Unless you test your plans, unforeseen circumstances, breakdowns, and actual effort cannot be determined.

Have a communication plan.
The communication plan not only establishes the protocol for communicating it should also define the roles for DRP execution.

  • Who will initiate the plan?
  • Do you have an elevation list that is maintained?
  • How will you communicate to stakeholders, the public, and customers?
  • Who establishes when recovery is complete?

The planning, disruptions, and effort to develop effective BU/DRP plans are significant.  As an investment, your BU/DRP efforts should be considered an insurance policy – you hope that you never have to use it, but glad that you have it when it is needed.

Discussion

View Comments for ““every web service’s worst nightmare””

Leave a comment

blog comments powered by Disqus