Gil, You oversimplify the argument.
If you overwrite the data once, with anything, it is gone. Forever and ever and ever. What remains is the possibility that the overwritten data can be ascertained from the residual data on the edges of the track that come about because the heads do not track over exactly the same position all the time. The theory is that if you overwrite the track enough times the head will have been "shaken" around enough that the extreme edges of the track have been overwritten. Theories abound as to how to use the track offset "noise" to recover old data, but as far as I can tell no-one has done this successfully unless the original contents were known. Techniques I've seen discussed over the years do not use equipment and software that will run efficiently on a BlackBerry, usually involve equipment that fits in a truck instead of a briefcase, and requires several 100 rereads just to come up with the possibility that something is there. Personally I think it is still something that belongs in the realm of sci-fi and spy thrillers, but we can't change the fact that some bureaucrats drank the cool aide and got religion on this. If this storage model does not have a shredding disk shredding feature then I'd suggest: 1) FDR/ERASE 2) TRKFRMT CYCLE(1) several times with thirty minute wait per execution. Several is whatever your statutory requirement needs 3) Degaussing 4) Sledgehammers RON > The debate rages. One side insists that one must overwrite the data many > times before it's really gone. For an opposing view, see: > > http://www.h-online.com/security/news/item/Secure-deletion-a-single- > overwrite-will-do-it-739699.html > > ... which mentions a more realistic concern that there may be residual data in > editor recovery files, page data sets, etc. And with modern virtual DASD with > technologies such as copy-on-write, you have little assurance that you're > overwriting the data you wish to secure. > > -- gil > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send email to > [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
