Tech Careers: Copy That and Snap It

Large and small businesses rely on high-tech companies to protect one of their most critical assets: information. These businesses will often purchase a disk array for this purpose.

If you find yourself interviewing for a high-tech position, you might get asked for your thoughts on how best to access a copy of a live, constantly changing piece of information.

Your answer (once you choose to read up on it!) should be a snap.

Snap copy technology is frequently used inside of a disk array in order to provide point-in-time copies of digital information. Note that snap copy technology can be used outside of a disk array as well. As we have discussed in previous posts, however, the features and functions of a disk array are important for you to know.

We’ve been using this simple diagram to illustrate the flow of information between an application and a persistent storage device. Consider the following diagram, where the application on the left (Application #1) has written information to a disk array.

 

Imagine that Application #2 wishes to run experiments on this data. Perhaps Application #2 is a financial application that wishes to create reports, modify the data, and run “what-if” scenarios. The problem here is that these experiments cannot disrupt or corrupt the data flow from Application #1.

Most disk arrays support the ability to make a “point-in-time copy”, “snapshot”, or “snap copy” of application data. Usually applications will “pause” just long enough to instruct the disk array to “split off” a copy of a certain piece of information. This copy can then be mounted and accessed by another application without corrupting the integrity of the production data. If application #1 pauses at 3:00 PM and instructs the disk array to create a snap copy, it can then continue modifying the data. The diagram below depicts Application #2 accessing a copy, while Application #1 has already continued to change the original data.

 

This is a very handy feature that can enable much more than just the simple running of experiments. Application #2 could be a backup application that safely stores the data in a remote location. It could be an application that performs a virus scan. Regardless of the capabilities of Application #2, Application #1 can rest assured that none of this activity will corrupt the primary copy.

Closely related to snap copy is a feature known as “continuous data protection”. CDP allows an application to access an old copy from any previous point in time(as opposed to a copy which was created at a specific point in time such as 3PM).

Knowing the details of these technologies is advantageous when conversing with someone in the high-tech industry. You can start with Wikipedia and read up on both CDP and Snapshot technologies. My corporation (EMC) has created a product (Symmetrix) which supports a feature known as Timefinder Business Continuance Volumes (BCVs). For more reading on BCVs I recommend this description.

Steve
http://stevetodd.typepad.com
Twitter: @SteveTodd
EMC Intrapreneur