On Thu, Feb 14, 2013 at 2:40 PM, Lizette Koehler 
<[email protected]<mailto:[email protected]>> wrote:
I have been asked by a friend to find out what is used to encrypt IMS Data
Bases that are PCI compliant.

The data is small.  But they just want to make sure no one can get to the
data.  If it goes to Tape I know what that should be, but the curiosity is
what is done on the data bases themselves.

Or do you just ensure the SAF rules are strong enough?

No, SAF cannot provide PCI compliance. You need strong encryption and/or 
tokenization on the PCI data (credit card numbers or, more specifically, the 
PANs - Primary Account Numbers - that are part of CCNs). It's a whole theology, 
and has its own peculiar rules, but the good news is that the PCI DSS document 
is quite readable and not that long.

One aspect is "PCI scope", which has to do with whether a given application or 
environment even has access to the plaintext. Using encryption/tokenization 
wisely can remove entire subsystems and even systems from PCI scope, which 
makes audits much faster and easier (and cheaper). For example, whole-database 
encryption may get you through a PCI audit-depending on the QSA (Qualified 
Security Assessor-the auditor), but every app that uses the data will be in 
scope.

In case you can't tell, this is one of the things we do here at Voltage 
Security, Inc., with our Voltage SecureData product, which includes native 
encryption and tokenization on z/OS. With application-level, Format-Preserving 
Encryption and Format-Preserving Tokenization, only those applications that 
will do decryption/detokenization are in scope-and with format-preserving data 
protection, that's not every application that uses the data.

Consider a credit card number: the only times you likely ever have the 
unprotected number are when it comes in (from a screen, partner, whatever-and 
if that source is "a web page", you encrypt the PAN *in the page*, so it's 
protected all the way through), when you send it on to the bank for settlement, 
and perhaps when fraud analysis finds a victim and needs to get back to the 
real data to report it. Otherwise, apps that "use" that number are likely to 
just be passing it along, perhaps checking that there *is* a value. So they 
don't need to do decryption-and since it's protected with a format-preserving 
technology, they don't even need any changes at all. Indexes still work, 
referential integrity is maintained-it's really reduces the pain of a data 
protection project significantly.

I've probably gone on enough here; if you'd like more information, drop me a 
line.

Cheers,
--
...phsiii

Phil Smith III
[email protected]<mailto:[email protected]>
Voltage Security, Inc.
www.voltage.com<http://www.voltage.com/>

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to