I found Martin's answer laugh-out-loud funny. But I think his recipe is far too limited. It's important to start at every bit boundary (not just word or byte boundaries), to employ a wide variety of variable word lengths, and to look at both big-endian and little-endian bit orders. I would also make sure to include all types of character encoding, including at least Unicode (UTF-xx), ASCII, and EBCDIC, plus several variants, such as code-shifted variants. Only then can you print out the results, and even that may not be enough.
:-) Seriously, I think there are some good suggestions. With respect to CPU, "no problem": this is the classic sort of one-time (or occasional) scan that could run at a lower service class and very successfully trade longer wall-clock time for lower CPU. The Y2K-inspired scanners and their progeny (examples: WebSphere Studio Asset Analyzer, Rational Asset Analyzer) would seem to be very good choices for sifting through application source code and looking for variables, column names, and other application-related artifacts that resemble credit card numbers. (And that could be very useful in an audit.) I tend to think these tools would be less adept at sifting through generic databases and files on a systemwide basis, but it's possible they would be a good match. Products like File Manager and File-AID seem somewhat relevant. However, I think I'd also look at something in the IBM Optim family. (In Optim parlance this is known as "e-discovery" functionality). I also think of products like DFSORT and Syncsort, which are tools that are good at parsing things -- a core competency for this type of inspection. SAFR might also be relevant, and it is very good at "needles in a gigantic haystack" sort of work, including making sense of even tape-based data. (SAFR is already being used for some of the more interesting, urgent, and gigantic audits in the global financial crisis.) If I understand a little more about the required task I could probably offer some more details and/or recommendations. - - - - - Timothy Sipples IBM Consulting Enterprise Software Architect Based in Tokyo, Serving IBM Japan / Asia-Pacific E-Mail: [email protected] ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
