Chapter 10. The Cone of Silence

The previous chapter looked at data protection from the disaster recovery perspective: how to ensure that mission critical data could be returned to an accessible state rapidly in the wake of an unplanned interruption. We looked at many products, topologies and solutions for ensuring the timely replication of data and its timely restoration to productive use following a disaster event.

We saw how many emerging technologies for enhancing data protection capitalize on multi- tier storage architectures that, in turn , take advantage of new storage interconnects such as switched fabrics (Fibre Channel) and true storage networking (iSCSI) protocols. This same infrastructure is enabling the implementation of new management paradigms like access frequency-based data migration that promise to improve the ways that we provision storage to applications.

What was left out of the previous discussion was the other aspect of disaster recovery, which is disaster avoidance . Disaster recovery planning seeks to eliminate those potential causes of unplanned data access interruption that can be eliminated and to minimize the impact of interruption potentials that cannot be eliminated.

Within the Disaster Recovery Planning context, disaster avoidance helps to minimize or eliminate risks that may lead to disasters. In many cases, it amounts to little more than the application of common sense practices such as the following:

• Identify and eliminate known disaster potentials by looking at facility physical security, sources of potential electrical or combustion hazards, airborne particulate counts (which can cause equipment damage and fire), water hazards, and so forth. Simple steps can make the workplace a safer environment.

• Prepare a detailed inventory of all equipment especially computer and communications components and make sure you keep it up-to-date. The inventory should cover all hardware, software, communications equipment, peripherals, backup media, and power conditioning equipment. The inventory should include the following information:

  • Model & Serial Numbers

  • Firmware revision levels

  • Quantity

  • Vendor/Supplier

  • Alternate Vendor/Supplier

  • Availability

• Examine network and storage interconnects for potential single points of failure and rectify these as needed. Redundant ports, redundant lines, redundant power supplies , etc. are all options that might have been ignored at the time of initial design and deployment and may be able to be added at minimal cost to reduce risk.

• Negotiate a standby agreement with a local dealer , reseller or integrator for priority replacement of hardware and other LAN components in case of an emergency or disaster.

• Evaluate your current backup strategy to determine its consistency with changes in business processes and applications.

• Test your backup media on a regular basis to ensure that you are able to recover your systems and databases from backup media.

This last point is especially important. Unfortunately, many computer users have a naive faith in computer storage media. Examples are replete of users entrusting their only copy of a critical file to single floppy disk or CD-ROM.

The simple fact is that all forms of magnetic (and optical) media degrade over time and will ultimately fail. There is far less certainty than is widely assumed as to when a tape, disk, or CD media will fail. Recent studies of CD recording media, for example, revealed that many brand name products did not live up to the resiliency claims of their manufacturers in accelerated aging tests.

In any event, the advice remains the same: Save your copies of your work REGULARLY and NOT to the same storage device that stores originals . Backup frequency will depend upon how much data you have and how regularly it changes. As a guideline, if you couldn't bear to lose something, make a copy. Choosing an alternative backup location is eminently sensible or your backup will suffer the same demise as the original.

In addition to disaster prevention through effective data replication, it is increasingly important to look at data disaster prevention from the perspective of security. This is a "new frontier" in storage technology, and one that is thrust upon us as a result of burgeoning networked storage techniques and architectures. This chapter will focus on the role of data security in a networked storage world.