i5/OS may be famous for security, but it doesn't go far enough in protecting your system's data; it doesn't differentiate how the data being accessed can be used. Because read/write/update authority is needed to use an interface such as a 5250 green-screen or a GUI to access the data, unnecessary authority for actions such as FTP can allow files to be moved off the system and intended securities to be bypassed. Therefore, it's wise to supplement your security with third-party solutions that can secure how data can be used with network applications.
One popular way of enhancing IBM's built-in security is to analyze requests—or "transactions"—by parsing SQL statements using a network security application. A manager can review a list of SQL statements and transactions to decide whether or not to allow the activity. This method provides increased security and effectively keeps users out of the system, but it's time-consuming and requires constant maintenance.
IBM uses a more elegant solution for its built-in security, treating requests as "objects" and eliminating the need to examine SQL statements. Many users don't realize that such functionality is available in third-party solutions as well, providing coverage of all the same exit points as popular transaction-based solutions. Let's look at how such a solution can save you time and money and improve your operations and security.
Lost at SQL (or How Transactions Washed the Manager Away)
Transaction-based solutions can export to Excel a list of requests so that a manager can review them and set actions to accept or reject. Sounds useful, right? Now imagine a spreadsheet with 700 lines and 13 columns, containing dense information about locations, servers, functions, transaction times, and jobs. Each line must be reviewed and its action set. Sure, you could get through those 700 lines, but I forgot to mention that they were all generated in three hours. How many will you get next time?
Now imagine that you approved the request Select * from Myfile. When you open your next spreadsheet, you find that a user attempted to access data using select * from myfile. Was it approved? No. This request was rejected because the words "select" and "myfile" were not capitalized. Your network security application is treating these as two different requests, when in fact they are the same. Therein lies the weakness of transaction-based security.
In reality, the problem is much greater than this. The number of SQL statements requiring review can grow exponentially, and it is impractical for a human operator to review them all, constantly updating the software to accept different permutations of the same request. That's why transaction-based security, although popular, is time-consuming, difficult to manage, and error-prone.
The Object of Thy Desire
Returning to the model that IBM has chosen, focusing on objects can relieve you of the hassles of maintaining permissions. By breaking transactions down into the objects being requested, object-based solutions can provide adaptive security that requires minimal input from operators and eliminates the need to parse SQL statements or to "memorize" every transaction.
Let's take another look at the request Select * from Myfile. When we tried to grant permission for this using transaction-based security, we had to backtrack a few hours later when a difference in case created a mismatch. We corrected this problem by also memorizing select * from myfile. Not mentioned above was a later request that came through as SELECT * FROM MYLIB/MYFILE. All wanted access to the same data, but all had to be memorized separately.
This problem disappears when you use object-based security. The object-based solution understands that the user wants "myfile," and it will allow proper access to this object regardless of the syntax. Gone are the bloated spreadsheets, the SQL statements, and the constant reviewing of actions. This means time saved for you, and it also reduces staffing costs and the possibility of configuration errors that can prevent legitimate users from doing their work.
Reclaim Your Time, Improve Your Security
An object-based approach gives you the power to enhance security, improve the user's experience, and put your time to better use. It's no wonder IBM went the object route when building security for the world's strongest operating system. If you're considering implementing third-party security for the first time, choosing an object-based solution is one of the best ways to increase your return on investment. And if you're already using third-party software that is transaction-based, the object model is definitely worth a look. A change could do you good.
Check out Bytware's offerings in the MC Showcase Buyer's Guide.
Mike Grant is CEO and head of development for Bytware, Inc.
LATEST COMMENTS
MC Press Online