Because they are separate rules, they should be separate remediations.
Of course if the upstream faillock BZ [1] was prioritized and dealt with sooner,
we probably wouldn't necessarily be having this discussion.


On Tue, Feb 12, 2019 at 10:45 AM Shawn Wells <> wrote:

On 2/12/19 11:49 AM, Marek Haicman wrote:
> Hello everyone,
> we have currently stumbled upon situation, where Ansible remediation
> snippet can either fix 3 different rules at once, or be very
> convoluted. Technical details aside [1] - what is your view of such
> approach?
> * Is it ok when remediation does change more than the rule that
> triggered it checks?

Current methodology ensures higher-level technologies can compose custom
security baselines (incl SCAP and remediation). Kind of like what SCAP
Workbench does.

If we can't track one configuration item to specific
XCCDF/OVAL/remediation, all that falls apart.

> * Do you prefer to have no remediation at all, to the remediation that
> does too much?
Would have to understand what "too much" means. Very surprised Ansible
wouldn't be able to remediate single configuration checks. Worst case
use the shell capabilities and run whatever the bash snippet would be.

> * Does answer to the questions above change between (--remediate)
> which is applied automatically, and bash roles or ansible playbooks,
> where you can check insides of the scripts and alter them before
> application?
If running --remediate, multiple CCEs are somehow grouped into a single
ansible action, how do I troubleshoot that?
scap-security-guide mailing list --
To unsubscribe send an email to
Fedora Code of Conduct:
List Guidelines:
List Archives: