Scanning docker container images

Looking for potential vulnerabilities in docker images is crucial before shipping these to customers or putting them into production. Scanning of images should be a part of any CI pipeline so that it’s ensured that shipped software is secured as possible and security vulnerabilities and CVEs are detected early in the process.

Docker engine comes in with 

that runs on Snyk engine to detect CVEs and vulnerabilities. A result of the scan will show also potential remediations (e.g. use newer base image). 

More details on: https://docs.docker.com/engine/scan/

Another solution is trivy from Aqua Security that serves a similar purpose https://github.com/aquasecurity/trivy. Running it after installation is as simple as 

openstack cli in a docker container

If you need to access your openstack cluster but there is no option to install packages on a jumpbox host that can access the cluster (lack of internet access or privileges) an alternative is to build locally a docker image that includes openstack CLI utility. Assumption is that the jumpbox host has docker installed and user can load docker images and run docker containers.

Firstly prepare a Dockerfile (based off python docker official images in this example):

Build docker image and provide as an argument path to stackrc file that includes details on how to access the cluster (endpoint, passwords etc).

After build is finished save docker image to tarball:

Copy tarball to a destination machine which can access openstack cluster and load docker image:

Run the container – you’ll and try listing e.g. servers