Testing DUKPT

As an acquirer, you can validate that your PIN translation command is working correctly even if you haven’t yet established connectivity to your Debit/EBT endpoint (or if you’ve established connectivity but don’t yet have your test key parts).

Typically when we’re pressed into this situation, this is what we do if we’re working with the Key-Up II HSM from Ian Donnelly Systems (‘IDS’):

For simplicity’s sake, we use a weak double-length key part.

We enter it as a single part (again, at this point, we want to keep things simple).
Typically, we use 0123456789ABCDEFFEDCBA9876543210.

We take the Cryptogram obtained out of that exercise and we use that as our Key Exchange Key (“KEK”) input.

Now, in my reading of the Thales manuals, performing this exercise with the Thales device is quite a bit more complicated.

[I’ll use Thales language here so we’re all on the same page…]

GOAL: What you need to test PIN translation is a ZPK encrypted under your LMK. [Refer to my previous post for details on the CI/CJ PIN transalation request/response.]

Note specifically where I describe “PIN Length” on the CJ command output I say:

“Although this field is not used to build the 0200 message formatted for the processor gateway, a value like ‘04’ or ‘05’ here is a pretty good indication that the translation occurred successfully.”

So, we can validate the workability of this command without a working endpoint or ‘real’ key parts from them, but – in order to do so – we need to construct a valid ZPK under LMK cryptogram. This value will allow you to create a completely valid CI command, assuming that:

You’ve got one or more valid BDK cryptograms loaded in your encryption database.

You can generate PIN-enabled Debit purchases from a test point-of-sale location.

You have addressibility from your test server to the Thales.

STEPS: My reading of the Thales manual is that you’ll need to execute the following steps (this is sort of ugly – numbers refer to specific sections in the Operations and Installation manual):

6.1 to create at least two 32-positions ZMK components (or, alternatively, 6.3 if you want to use components of your own chosing).

6.4 to get a 32-position ZMK under LMK.

7.1 to generate random ZPK (uses 6.4 ZMK under LMK output as input here) and encrypt under ZMK and LMK. The goal is that second cryptogram that comes back, the one oddly named “ZPK encrypted for bank” in the Thales document. That’s the ZPK under the LMK that you need for your tests.

Obtaining this information should allow you to validate the critical part of your encryption infrastructure prior to getting connectivity to your Debit/EBT gateway partner. Of course, once formal inter-system testing is established you still need to transport the encrypted PIN block in an 0200 request to your endpoint. But doing the exercise I describe above here first will allow you to eliminate any self-doubt you may have about your internal PIN translation activities prior to getting an external party involved in the testing


What are your thoughts?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s