Secret leaks often occur when a sensitive piece of authentication data is stored with the source code of an application. Considering the source code is intended to be deployed across multiple assets, including source code repositories or application hosting servers, the secrets might get exposed to an unintended audience.
In most cases, trust boundaries are violated when a secret is exposed in a source code repository or an uncontrolled deployment environment. Unintended people who don’t need to know the secret might get access to it. They might then be able to use it to gain unwanted access to associated services or resources.
The trust issue can be more or less severe depending on the people’s role and entitlement.
In Dockerfiles, hard-coded secrets and secrets passed through as variables or created at build-time will cause security risks. The secret information can be exposed either via the container environment, the image metadata, or the build environment logs.
The consequences vary greatly depending on the situation and the secret-exposed audience. Still, two main scenarios should be considered.
Financial losses can occur when a secret is used to access a paid third-party-provided service and is disclosed as part of the source code of client applications. Having the secret, each user of the application will be able to use it without limit to use the third party service to their own need, including in a way that was not expected.
This additional use of the secret will lead to added costs with the service provider.
Moreover, when rate or volume limiting is set up on the provider side, this additional use can prevent the regular operation of the affected application. This might result in a partial denial of service for all the application’s users.
A downgrade can happen when the disclosed secret is used to protect security-sensitive assets or features of the application. Depending on the affected asset or feature, the practical impact can range from a sensitive information leak to a complete takeover of the application, its hosting server or another linked component.
For example, an application that would disclose a secret used to sign user authentication tokens would be at risk of user identity impersonation. An attacker accessing the leaked secret could sign session tokens for arbitrary users and take over their privileges and entitlements.
Best practices recommend using a secret vault for all secrets that must be accessed at container runtime. This will ensure the secret’s security and prevent any further unexpected disclosure. Depending on the development platform and the leaked secret type, multiple solutions are currently available.
For all secrets that must be accessed at image build time, it is recommended to rely on Docker Buildkit’s secret mount options. This will prevent secrets from being disclosed in image’s metadata and build logs.
Additionally, investigations and remediation actions should be conducted to ensure the current and future security of the infrastructure.
Revoke the secret
Revoke any leaked secrets and remove them from the application source code.
Before revoking the secret, ensure that no other applications or processes are using it. Other usages of the secret will also be impacted when the secret is revoked.
Analyze recent secret use
When available, analyze authentication logs to identify any unintended or malicious use of the secret since its disclosure date. Doing this will allow determining if an attacker took advantage of the leaked secret and to what extent.
This operation should be part of a global incident response process.
The following code sample generates a new SSH private key that will be stored in the generated image. This key should be considered as compromised. Moreover, the SSH key encryption passphrase is also hardcoded.
FROM example # Noncompliant RUN ssh-keygen -N "passphrase" -t rsa -b 2048 -f /etc/ssh/rsa_key RUN /example.sh --ssh /etc/ssh/rsa_key
The following code sample uses a seemingly hidden password which is actually leaked in the image metadata after the build.
FROM example ARG PASSWORD # Noncompliant RUN wget --user=guest --password="$PASSWORD" https://example.com
FROM example
RUN --mount=type=secret,id=ssh,target=/etc/ssh/rsa_key \
/example.sh --ssh /etc/ssh/rsa_key
FROM example
RUN --mount=type=secret,id=wget,target=/home/user/.wgetrc \
wget --user=guest https://example.com
For runtime secrets, best practices recommend relying on a vault service to pass secret information to the containers. Docker environment provides Swarm services that implement such a feature.
If such an option can not be considered, store the runtime secrets in an environment file
such as .env and then start the container with the --env-file argument:
docker run --env-file .env myImage
It is then important to ensure that the environment files are securely stored and generated.