Posted by Cesar Cerrudo
Have you ever entered an office and seen a pile of money sitting unattended and easily accessible on a desk? How many people in your company have a key or combination to a safe with money inside and can open that safe without any controls? Do you leave money in a non-secure place that everyone knows about and can freely access?
Your probable answer to all these questions is NO, which makes sense—what doesn't make sense is how so many companies don't think the same way about data. I think data is worth a lot of money if you consider how important it is in terms of cost to the company: cost when it's stolen, cost when it's not available, et cetera. Data deserves to be protected as if it were money, but most of it is freely available by way of corporate databases; once you access the database you can play with the data at will, bypassing only modest controls or restrictions.
Of course you need a username and password to make the initial connection or exploit a SQL injection vulnerability (for example), but we all know it's not that difficult to get access, as shown by recent events. A lot of user passwords have been leaked, obtained from successfully-hacked companies that apparently didn't protect their data properly and, as a result, put their business at serious risk.
The main cause of data breaches is an improperly-secured database. Unfortunately, when it comes to database security, most companies are ages away from doing it properly.
For example, if the statement used to access a table is always:
Select * from user_accounts where user_email = X
why would you let anyone execute the next SQL statement:
Select * from user_accounts
Why not use stored procedures exclusively and remove all direct access to tables? Why not set alerts to trigger when common SQL injection-related errors occur? Why not monitor the database in real time to detect suspicious activities? Why not create a table named "important_data_is_here" and fire all the alarms when someone tries to access it?
Database servers don’t have advanced security features, but there are numerous third-party solutions that do—Database Activity Monitoring (DAM: a kind of database IPS) being one example—and a very small percentage of companies are using them. This isn't just about using a DAM product (which won't solve all your problems anyway), but it does provide a good start, allowing you to know in “real time” whether someone is digging around and playing with your databases.
If you don’t want to spend money on third-party solutions, there are ways to do customized DAM builds using database-provided functionality that involves creating alerts on specific actions, setting custom permissions on different database objects, monitoring and analyzing logs, creating a database honeypot and querying and analyzing system tables; however, when you have dozens (or hundreds) of databases, the difficulty level increases quickly and you'll need to have database security-wise personnel in place.
I've researched database security for more than 10 years and have seen database software vendors seriously improve the security of their products, but I have yet to see noticeable improvements at the company level—as indicated by all the breaches we know and don’t know about—corporations still don't seem to get that data is money.