Sound IT risk management is all about identifying critical data assets and giving them the most protection. The more critical an asset, the more defenses should be around it. Unfortunately, when it comes to databases, most companies get that formula backward.
The problem is that database performance can take priority over security at many companies. Rather than balancing security and performance issues, database security is too often left for some other time.
"DBAs and the application developers just don't have time or don't want to deal with security. It increases the cost of their product development," says Julie Lockner, an analyst at research firm Enterprise Strategy Group. They're being asked to add more applications and features, and deal with rising data volume, and that's making their test cycles longer. Says Lockner: "It's a priority thing: Do we get the features out? Or do we take the extra cycles to tie in and add the security layers around it?"
Malicious insiders and wily hackers can take advantage of this priority war within IT departments. They're accessing data they shouldn't, launching SQL injection attacks to take advantage of poorly protected app-to-database links, and exploiting vulnerabilities in database management systems to get into potentially huge and valuable data stores.
The only way to truly protect data is to make critical database security a top concern. It starts with these three principles of database protection.
Many companies aren't able to protect mission-critical data because they simply don't understand how all the moving parts of their database environments work. For controls to work, IT must have a clear understanding of where the important data is, who's using it, and how it's being used.
"You have one data store, but you might have many applications hooked into it. You might not know who it is that's using the systems if you've given out a lot of privileges," says Mel Shakir, CTO of NitroSecurity, a database activity monitoring (DAM) and security information and event monitoring company recently purchased by McAfee. "And you might not even know where the critical data is if it's been copied off the system and moved to, say, test databases somewhere else."
Valuable steps include scanning for unsanctioned, rogue databases that might have been set up on the fly by other departments, documenting privilege schemas, and classifying a company's database assets by risk according to the type of data they hold. That can help get more out of database security investments.
Once IT teams know where all your databases are, they can make sure they're securely configured and patched, and use vulnerability assessment to decide what level of protection they need. For example, they can decide if they warrant constant oversight through activity-monitoring software to track what users are doing in these data stores at all times.
Download the Oct. 31, 2011 InformationWeek digital supplement