[Open-scap] Using profiles not distributed in

[Shawn Wells]

On 2/8/19 2:34 PM, Greg Silverman wrote:
> Let me ask in a different way.
>
> DISA published xml files withhttps://iasecontent.disa.mil/stigs/zip/U_Red_Hat_Enterprise_Linux_7_V2R2_STIG_SCAP_1-2_Benchmark.zip. The zip's xml file contains a list of vulnerabilities for RHEL7, the Version 2 Release 2 (V2R2) selection of vulnerabilities. scap-security-guides versions 1.40+ contain a DISA profile and that profile contains the V1R4 list of vulnerabilities.
>
> 1. Can oscap v 1.2.17 consume the xml files at the DISA URL and evaluate a RHEL7 machine?


DISA only publishes what's called XCCDF -- essentially, human-readable 
prose. DISA does not publish any automation that would result in a 
pass/fail configuration scan.

The most any SCAP tool could do with this content, including OpenSCAP, 
would be to transform it from XML to HTML to ease reading:

$ oscap xccdf generate guide \
U_Red_Hat_Enterprise_Linux_7_V2R2_STIG_SCAP_1-2_Benchmark.xml \
 > ~/disa-guide.html


> 2. How do xml files like the ones at that URL get incorporated in a scap-security-guide, as was done with the DISA V1R4 files?

Manually.

Unfortunately DISA does not coordinate their content with DoD, NIST, 
NSA, or even Red Hat. These parties only find out about DISA's content 
when it's made publicly available.

And also unfortunately, DISA does not provide a changelog of what was 
changed. That means someone needs to go through the DISA content and 
compare it by hand. From there a series of tickets are opened to discuss 
alignment:

https://github.com/ComplianceAsCode/content/issues?q=is%3Aissue+is%3Aopen+label%3A%22DISA+Content+Issues%22

Once that ticket queue is resolved, the two bodies of content will be in 
alignment.