OpenSCAP-daemon

Continuously evaluate your infrastructure for SCAP compliance! Avoid copying big SCAP files around, avoid having to type long IDs, avoid writing ad-hoc bash scripts to solve your compliance needs!

Project Description

OpenSCAP-daemon is a service that performs SCAP scans of bare-metal machines, virtual machines and containers. These scans can be either one-shot or continuous according to a schedule. You can interact with the service using the provided oscapd-cli tool or via the DBus interface.

Motivation

The OpenSCAP project has progressed greatly over the past years and now provides very nice tooling to perform solicited one-off SCAP evaluation of the machine it runs on. Unsolicited, continuous or planned evaluation has always been out of scope of OpenSCAP to avoid feature creep. The previously mentioned use-case is very desirable and has been requested many times. We feel that now the time is right to start a project that helps you run oscap and does evaluation for you. OpenSCAP-daemon is such a project.

The project currently comprises of two parts, the daemon that runs in the background sleeping until a task needs processing, and the command-line tool that talks to the aforementioned daemon using dbus. Do not be alarmed, the command-line tool is much easier to use than pure oscap for common use-cases.

Features

• SCAP evaluation of the following assets using OpenSCAP -- a NIST-certified scanner
  • local machine -- oscap
  • remote machine -- oscap-ssh
  • virtual machine -- oscap-vm
  • container -- oscap-docker
• flexible task definition and planning
  • use any valid SCAP content -- for example SCAP Security Guide, NIST USGCB, or even RHSA OVAL
  • evaluate daily, weekly, monthly or in custom intervals
  • evaluate on demand
• parallel task processing
• results storage -- query ARFs of past results, generate HTML reports, get oscap stdout/stderr and exit codes
• command-line interface
• dbus API
• fully automated CVE evaluation of containers using OpenSCAP and Atomic.mount
• Cockpit integration (planned)

Key Goals & Design Decisions

We have learned many important lessons when developing the lower layers of the SCAP evaluation stack that we want to address in this project.

• useful defaults -- just pressing Enter and not providing any details should still yield a valid setup
• simplicity -- we avoid RDBMS and instead use features of the filesystem
• datastreams -- SDS (source datastream) and ARF (results datastream) are both used as primary data formats for maximum compatibility between various tools
• interactive CLI -- the CLI should be as interactive as possible, user shouldn't need to type any IDs or other lengthy options

Example Use-Cases

Scan a container or container image on Atomic Host

Atomic host can use the functionality in OpenSCAP-Daemon to perform vulnerability scans of containers and container images using the atomic scan command.

To use this functionality, install atomic. Then install openscap-daemon either in standalone mode or as a SPC container image. When the daemon is running the atomic scan functionality is available.

Scan all containers or all container images on Atomic Host

The atomic scan command has command-line arguments --images, --containers and --all that scan all images, all container and everything respectively.

Scan local machine every day at 1:00 AM UTC

OpenSCAP-daemon thinks in terms of tasks. Let us first define the task we want to perform:

# interactively create a new task
oscapd-cli task-create -i
Creating new task in interactive mode
Title: Daily USGCB
Target (empty for localhost): 
Found the following SCAP Security Guide content: 
        1:  /usr/share/xml/scap/ssg/content/ssg-fedora-ds.xml
        2:  /usr/share/xml/scap/ssg/content/ssg-firefox-ds.xml
        3:  /usr/share/xml/scap/ssg/content/ssg-java-ds.xml
        4:  /usr/share/xml/scap/ssg/content/ssg-rhel6-ds.xml
        5:  /usr/share/xml/scap/ssg/content/ssg-rhel7-ds.xml
Choose SSG content by number (empty for custom content): 4
Tailoring file (absolute path, empty for no tailoring): 
Found the following possible profiles:
        1:  CSCF RHEL6 MLS Core Baseline (id='xccdf_org.ssgproject.content_profile_CSCF-RHEL6-MLS')
        2:  United States Government Configuration Baseline (USGCB) (id='xccdf_org.ssgproject.content_profile_usgcb-rhel6-server')
        3:  Common Profile for General-Purpose Systems (id='xccdf_org.ssgproject.content_profile_common')
        4:  PCI-DSS v3 Control Baseline for Red Hat Enterprise Linux 6 (id='xccdf_org.ssgproject.content_profile_pci-dss')
        5:  Example Server Profile (id='xccdf_org.ssgproject.content_profile_CS2')
        6:  C2S for Red Hat Enterprise Linux 6 (id='xccdf_org.ssgproject.content_profile_C2S')
        7:  Common Profile for General-Purpose SystemsUpstream STIG for RHEL 6 Server (id='xccdf_org.ssgproject.content_profile_stig-rhel6-server-upstream')
        8:  Common Profile for General-Purpose SystemsServer Baseline (id='xccdf_org.ssgproject.content_profile_server')
        9:  Red Hat Corporate Profile for Certified Cloud Providers (RH CCP) (id='xccdf_org.ssgproject.content_profile_rht-ccp')
Choose profile by number (empty for (default) profile): 2
Online remediation (1, y or Y for yes, else no):                                                                   
Schedule:                                                                   
 - not before (YYYY-MM-DD HH:MM in UTC, empty for NOW): 2014-07-30 01:00
 - repeat after (hours or @daily, @weekly, @monthly, empty or 0 for no repeat): @daily
Task created with ID '1'. It is currently set as disabled. You can enable it with `oscapd-cli task 1 enable`.

As the command-line interface suggests, we need to enable the task.

# enable previously created task of given ID
oscapd-cli task 1 enable

We may also want to see the HTML guide of our specified task to confirm it will do what we need.

# get the HTML guide of task of ID 1
oscapd-cli task 1 guide > guide.html
# open the guide in firefox
firefox guide.html

At this point oscapd will evaluate the local machine at 1:00 AM UTC every day and store all the results. To finish this use-case, lets see how we can query the results after a week of evaluations.

# list all available results of task 1
$ oscapd-cli result 1
7
6
5
4
3
2
1
# get the verbatim results ARF of the 4th result of task 1
oscapd-cli result 1 4 arf > exported-arf.xml
# get the HTML report of previously mentioned result
oscapd-cli result 1 4 report > report.html
# open the report in firefox
firefox report.html

Solicited evaluation

Sometimes we may want to run the evaluation outside the schedule for testing or other purposes. The task may even be scheduled to never run automatically! Such tasks are sometimes necessary.

# run task of ID 1 immediately
oscapd-cli task 1 run
# query available results
oscapd-cli result 1
8
7
6
# [snip]
# fetch ARF of result 8 of task 1
oscapd-cli result 1 8 arf > exported-arf.xml

Evaluate something else than local machine

Every task has a target attribute that can take various forms:

• localhost -- scan the local machine, the same machine the daemon runs on
• ssh://[email protected] -- scan remote machine of given IP with given username
  • make sure you can log onto the same machine non-interactively!
• ssh+sudo://[email protected] -- scan remote machine of given IP with given username with sudo privileges
  • sudo mustn't require tty
• vm://qemu+kvm://localhost/VM1 -- virtual machine -- work in progress, subject to change
• docker://container_id -- local container -- work in progress, subject to change

The rest of the use-case is similar to previously mentioned use-cases. It is important to remark that the SCAP content only needs to be available on the local machine -- the machine that runs OpenSCAP-daemon. It is not necessary to perform any extra manual action to get the content to the scanned machines, this is done automatically.

Scan all images in my registry to make sure no vulnerable images are published

When maintaining a registry it makes sense to unpublish images that have known vulnerabilities to prevent people from using them.

We need to react to the CVE feeds changing and re-scan the images and of course we need to scan all new images incoming into the registry.

This is a future use-case that hasn't been fully implemented yet.

Requirements

• python2 >= 2.6 OR python3 >= 3.2
  • full source compatibility with python2 and python3
• OpenSCAP >= 1.2.6
• dbus-python
• flask >= 1.0.2
• (optional) Atomic >= 1.4
• (optional) docker

Running the test-suite

The test-suite can be run without installing the software.

cd openscap-daemon
cd tests
./make_check

Installation on Linux (standalone on host)

cd openscap-daemon
# as a python2 application
sudo python2 setup.py install
# as a python3 application
sudo python3 setup.py install

Building a container with OpenSCAP Daemon

Containerized version of OpenSCAP Daemon is used as a backend for the 'atomic scan' command. Atomic scan can scan containers and images for vulnerabilities and configuration compliance.

You can build and install the container image using these commands:

./generate-dockerfile.py
docker build -t openscap .
atomic install openscap

At this point you can run 'atomic scan' on the host. The image is not meant to be run outside of the atomic command.

The image is based on Fedora and contains OpenSCAP, OpenSCAP Daemon and SCAP Security Guide as they are available in Fedora packages. To install your local working tree of OpenSCAP Daemon instead, add --daemon-from-local to the ./generate-dockerfile.py. If you need the latest code from upstream git of OpenSCAP and/or SCAP Security Guide instead, pass --openscap-from-git and/or --ssg-from-git to the ./generate-dockerfile.py.

REST API

REST API Endpoints

REST API Examples

Get all tasks:

curl -i http://127.0.0.1:5000/tasks/ -X GET

Create a new task:

newtask.json 
{
    "taskTitle":"New Task test",
    "taskTarget":"localhost",
    "taskSSG":"/usr/share/xml/scap/ssg/content/ssg-jre-ds.xml",
    "taskTailoring":"",
    "taskProfileId":"xccdf_org.ssgproject.content_profile_stig-java-upstream",
    "taskOnlineRemediation":"1",
    "taskScheduleNotBefore":"",
    "taskScheduleRepeatAfter":""
}
  
curl -i http://127.0.0.1:5000/tasks/ -X POST -H "Content-Type: application/json" -d '@newtask.json'

Update an existing task:

updatetask.json
{
    "taskTitle":"New Task test modified",
    "taskTarget":"localhost",
    "taskSSG":"",
    "taskTailoring":"",
    "taskProfileId":"",
    "taskOnlineRemediation":"yes",
    "taskScheduleNotBefore":"",
    "taskScheduleRepeatAfter":""
}
    
curl -i http://127.0.0.1:5000/tasks/1/ -X PUT -H "Content-Type: application/json" -d '@updatetask.json'

Get info from existing task:

curl -i http://127.0.0.1:5000/tasks/1/ -X GET

Get guide info from existing task:

curl -i http://127.0.0.1:5000/tasks/1/guide/ -X GET

Get result from existing task:

curl -i http://127.0.0.1:5000/tasks/1/result/1/ -X GET

Delete all results from existing task:

curl -i http://127.0.0.1:5000/tasks/1/result/ -X DELETE

Delete result from existing task:

curl -i http://127.0.0.1:5000/tasks/1/result/1/ -X DELETE

Launch existing task:

curl -i http://127.0.0.1:5000/tasks/1/run/ -X GET

Enable/Disable existing task:

curl -i http://127.0.0.1:5000/tasks/1/enable/ -X PUT
curl -i http://127.0.0.1:5000/tasks/1/disable/ -X PUT

Get all SSGs installed on the system:

curl -i http://127.0.0.1:5000/ssgs/ -X GET

Get SSG's profiles using data from a JSON request:

ssg.json
{
    "ssgFile": "/usr/share/xml/scap/ssg/content/ssg-centos7-ds.xml",
    "tailoringFile": ""
}
    
curl -i http://127.0.0.1:5000/ssgs/ -X POST -H "Content-Type: application/json" -d '@ssg.json'

API Consumers

Please do not rely on the API just yet, we reserve the right to make breaking changes. The API will stabilize in time for 1.0.0 release.

OpenSCAP-daemon provides a stable dbus API that is designed to be used by other projects.

Atomic Integration

OpenSCAP-daemon is used to implement the atomic scan functionality. atomic scan allows users to scan containers and container images for vulnerabilities.

Cockpit Integration

Features:

• declare new tasks, schedule when they run, set how they repeat
• generate HTML guides of scheduled tasks
• show past results of tasks
• get ARFs, HTML reports for past results
• set tasks to automatically push results to external result stores

Foreman Integration

Provide a way to reliably do one-off tasks. Unify various oscap runners into one code-base.