We’ve looked so far at data backup, the automation thereof, and the recoverability that’s essential. This week, consider a quick, broader look at infrastructure. If the lights went out, how would you operate? Folks in the Northeast are thinking about this, and even in the Midwest we’re all too familiar with ice storms and their power grid impacts.

It’s handy that just the other day, at an i2E event near us in Tulsa, Oklahoma, I met a pair of entrepreneurs from a “bunker” style vault data center, Titan Private Security Vault. They already have stories of millions of dollars saved when their clients–rather large enterprises in the region–were able to shift operations on a moment’s notice to a secondary location (TPSV) when the grid went dark. In one case, it was due to a major disruption in downtown Tulsa. In another case, it was a massive ice storm that made headlines across the nation.

Point is, these businesses had a plan in place that included secondary, leased workstations and infrastructure, allowing them to get payroll completed and handle other mission critical functions even when the rest of the city was dark.

It may be that this level of redundancy is beyond your budget. It may not be. Either way, thinking ahead to how you would keep a heartbeat going if the whole state lost electricity is a key part of a disaster recovery plan, or at least a “disaster aversion plan.”

This week, we consider the question, “If a tree falls in the forest, and nobody is there to hear it, does it make a sound?” In the DR world, it goes like this: “If a data backup is taken and nobody tests its recoverability/validity, is it worth more than a sour grape?” All too often, we rely on “having a backup” but we have never tested our recovery scenarios. If our backups haven’t been tested, guess what? Disaster, waiting to happen.

A person close to me was once fired by a major corporation because an email server crashed, and when a backup restoration was executed, it was only then determined that the proprietary email server database, which shall remain nameless, wasn’t backed up at all. All email: gone. My associate: gone (fired). Career Limiting Event.

Now in this case, there was some trouble with the agent that was supposed to allow the hot database file to be backed up, and actually the whole scenario was in place before my associate arrived. It didn’t matter, though–thousands upon thousands of emails lost, and nobody realized that these critical files we’re being backed up properly. “We have a backup” turned out to be wrong.

Test. Practice a recovery. If a tree falls…