Static Application Security Testing (SAST) (FREE)

All open source (OSS) analyzers were moved from GitLab Ultimate to GitLab Free in GitLab 13.3.

NOTE: The whitepaper "A Seismic Shift in Application Security" explains how 4 of the top 6 attacks were application based. Download it to learn how to protect your organization.

If you’re using GitLab CI/CD, you can use Static Application Security Testing (SAST) to check your source code for known vulnerabilities. You can run SAST analyzers in any GitLab tier. The analyzers output JSON-formatted reports as job artifacts.

With GitLab Ultimate, SAST results are also processed so you can:

• See them in merge requests.
• Use them in approval workflows.
• Review them in the security dashboard.

For more details, see the Summary of features per tier.

The results are sorted by the priority of the vulnerability:

1. Critical
2. High
3. Medium
4. Low
5. Info
6. Unknown

A pipeline consists of multiple jobs, including SAST and DAST scanning. If any job fails to finish for any reason, the security dashboard does not show SAST scanner output. For example, if the SAST job finishes but the DAST job fails, the security dashboard does not show SAST results. On failure, the analyzer outputs an exit code.

Use cases

• Your code has a potentially dangerous attribute in a class, or unsafe code that can lead to unintended code execution.
• Your application is vulnerable to cross-site scripting (XSS) attacks that can be leveraged to unauthorized access to session data.

Requirements

SAST runs in the test stage, which is available by default. If you redefine the stages in the .gitlab-ci.yml file, the test stage is required.

To run SAST jobs, by default, you need GitLab Runner with the docker or kubernetes executor. If you're using the shared runners on GitLab.com, this is enabled by default.

WARNING: Our SAST jobs require a Linux/amd64 container type. Windows containers are not yet supported.

WARNING: If you use your own runners, make sure the Docker version installed is not 19.03.0. See troubleshooting information for details.

Supported languages and frameworks

GitLab SAST supports a variety of languages, package managers, and frameworks. Our SAST security scanners also feature automatic language detection which works even for mixed-language projects. If any supported language is detected in project source code we automatically run the appropriate SAST analyzers.

You can also view our language roadmap and request other language support by opening an issue.

Note that the Java analyzers can also be used for variants like the Gradle wrapper, Grails, and the Maven wrapper.

Multi-project support

Introduced in GitLab 13.7.

GitLab SAST can scan repositories that contain multiple projects.

The following analyzers have multi-project support:

• Bandit
• ESLint
• Gosec
• Kubesec
• NodeJsScan
• MobSF
• PMD
• Security Code Scan
• Semgrep
• SpotBugs
• Sobelow

Enable multi-project support for Security Code Scan

Multi-project support in the Security Code Scan requires a Solution (.sln) file in the root of the repository. For details on the Solution format, see the Microsoft reference Solution (.sln) file.

Supported distributions

The default scanner images are build off a base Alpine image for size and maintainability.

FIPS-enabled images

Introduced in GitLab 14.10.

GitLab offers Red Hat UBI versions of the images that are FIPS-enabled. To use the FIPS-enabled images, you can either:

• Set the SAST_IMAGE_SUFFIX to -fips.
• Add the -fips extension to the default image name.

For example:

variables:
  SAST_IMAGE_SUFFIX: '-fips'

include:
  - template: Security/SAST.gitlab-ci.yml

Making SAST analyzers available to all GitLab tiers

All open source (OSS) analyzers have been moved to the GitLab Free tier as of GitLab 13.3.

Summary of features per tier

Different features are available in different GitLab tiers, as shown in the following table:

Contribute your scanner

The Security Scanner Integration documentation explains how to integrate other security scanners into GitLab.

Configuration

To configure SAST for a project you can:

• Use Auto SAST, provided by Auto DevOps.
• Configure SAST manually.
• Configure SAST using the UI (introduced in GitLab 13.3).

Configure SAST manually

To enable SAST you must include the SAST.gitlab-ci.yml template provided as a part of your GitLab installation.

Add the following to your .gitlab-ci.yml file:

include:
  - template: Security/SAST.gitlab-ci.yml

The included template creates SAST jobs in your CI/CD pipeline and scans your project's source code for possible vulnerabilities.

The results are saved as a SAST report artifact that you can later download and analyze. Due to implementation limitations, we always take the latest SAST artifact available.

Configure SAST in the UI

You can enable and configure SAST in the UI, either with default settings, or with customizations. The method you can use depends on your GitLab license tier.

• Configure SAST in the UI with default settings.
• Configure SAST in the UI with customizations. (ULTIMATE)

Configure SAST in the UI with default settings

Introduced in GitLab 13.9

NOTE: The configuration tool works best with no existing .gitlab-ci.yml file, or with a minimal configuration file. If you have a complex GitLab configuration file it may not be parsed successfully, and an error may occur.

To enable and configure SAST with default settings:

1. On the top bar, select Menu > Projects and find your project.
2. On the left sidebar, select Security & Compliance > Configuration.
3. In the SAST section, select Configure with a merge request.
4. Review and merge the merge request to enable SAST.

Pipelines now include a SAST job.

Configure SAST in the UI with customizations (ULTIMATE)

• Introduced in GitLab 13.3.
• Improved in GitLab 13.4.
• Improved in GitLab 13.5.

NOTE: The configuration tool works best with no existing .gitlab-ci.yml file, or with a minimal configuration file. If you have a complex GitLab configuration file it may not be parsed successfully, and an error may occur.

To enable and configure SAST with customizations:

1. On the top bar, select Menu > Projects and find your project.

2. On the left sidebar, select Security & Compliance > Configuration.

3. If the project does not have a .gitlab-ci.yml file, select Enable SAST in the Static Application Security Testing (SAST) row, otherwise select Configure SAST.

4. Enter the custom SAST values.

   Custom values are stored in the .gitlab-ci.yml file. For CI/CD variables not in the SAST Configuration page, their values are inherited from the GitLab SAST template.

5. Optionally, expand the SAST analyzers section, select individual SAST analyzers and enter custom analyzer values.

6. Select Create Merge Request.

7. Review and merge the merge request.

Pipelines now include a SAST job.

Overriding SAST jobs

WARNING: Beginning in GitLab 13.0, the use of only and except is no longer supported. When overriding the template, you must use rules instead.

To override a job definition, (for example, change properties like variables or dependencies), declare a job with the same name as the SAST job to override. Place this new job after the template inclusion and specify any additional keys under it. For example, this enables FAIL_NEVER for the spotbugs analyzer:

include:
  - template: Security/SAST.gitlab-ci.yml

spotbugs-sast:
  variables:
    FAIL_NEVER: 1

Pinning to minor image version

While our templates use MAJOR version pinning to always ensure the latest analyzer versions are pulled, there are certain cases where it can be beneficial to pin an analyzer to a specific release. To do so, override the SAST_ANALYZER_IMAGE_TAG CI/CD variable in the job template directly.

In the example below, we pin to a minor version of the semgrep analyzer and a specific patch version of the brakeman analyzer:

include:
  - template: Security/SAST.gitlab-ci.yml

semgrep-sast:
  variables:
    SAST_ANALYZER_IMAGE_TAG: "2.16"

brakeman-sast:
  variables:
    SAST_ANALYZER_IMAGE_TAG: "2.21.1"

Customize rulesets (ULTIMATE)

• Introduced in GitLab 13.5.
• Added support for passthrough chains. Expanded to include additional passthrough types of file, git, and url in GitLab 14.6.
• Added support for overriding rules in GitLab 14.8.

You can customize the default scanning rules provided by our SAST analyzers. Ruleset customization supports the following that can be used simultaneously:

• Disabling predefined rules. Available for all analyzers.
• Overriding predefined rules. Available for all analyzers.
• Modifying the default behavior of a given analyzer by synthesizing and passing a custom configuration. Available for only nodejs-scan, gosec, and semgrep.

To customize the default scanning rules, create a file containing custom rules. These rules are passed through to the analyzer's underlying scanner tools.

To create a custom ruleset:

1. Create a .gitlab directory at the root of your project, if one doesn't already exist.
2. Create a custom ruleset file named sast-ruleset.toml in the .gitlab directory.

Disable predefined analyzer rules

To disable analyzer rules:

1. Set the disabled flag to true in the context of a ruleset section

2. In one or more ruleset.identifier sub sections, list the rules that you want disabled. Every ruleset.identifier section has:

• a type field, to name the predefined rule identifier that the targeted analyzer uses.
• a value field, to name the rule to be disabled.

Example: Disable predefined rules of SAST analyzers

In the following example, the disabled rules are assigned to eslint and sobelow by matching the type and value of identifiers:

[eslint]
  [[eslint.ruleset]]
    disable = true
    [eslint.ruleset.identifier]
      type = "eslint_rule_id"
      value = "security/detect-object-injection"

  [[eslint.ruleset]]
    disable = true
    [eslint.ruleset.identifier]
      type = "cwe"
      value = "185"

[sobelow]
  [[sobelow.ruleset]]
    disable = true
    [sobelow.ruleset.identifier]
      type = "sobelow_rule_id"
      value = "sql_injection"

Those vulnerabilities containing the provided type and value are now disabled, meaning they won't be displayed in Merge Request nor the Vulnerability Report.

Override predefined analyzer rules

To override analyzer rules:

1. In one or more ruleset.identifier subsections, list the rules that you want to override. Every ruleset.identifier section has:

   • a type field, to name the predefined rule identifier that the targeted analyzer uses.
   • a value field, to name the rule to be overridden.

2. In the ruleset.override context of a ruleset section, provide the keys to override. Any combination of keys can be overridden. Valid keys are:

   • description
   • message
   • name
   • severity (valid options are: Critical, High, Medium, Low, Unknown, Info)

Example: Override predefined rules of SAST analyzers

Before adding a ruleset, we verify which vulnerability will be overwritten by viewing the gl-sast-report.json:

"identifiers": [
        {
          "type": "gosec_rule_id",
          "name": "Gosec Rule ID G307",
          "value": "G307"
        },
        {
          "type": "CWE",
          "name": "CWE-703",
          "value": "703",
          "url": "https://cwe.mitre.org/data/definitions/703.html"
        }
      ]

In the following example, rules from gosec are matched by the type and value of identifiers and then overridden:

[gosec]
  [[gosec.ruleset]]
    [gosec.ruleset.identifier]
        type = "CWE"
        value = "703"
    [gosec.ruleset.override]
      severity = "Critical"

If a vulnerability is found with a type CWE with a value of 703 then the vulnerability severity is overwritten to Critical.

Synthesize a custom configuration

To create a custom configuration, you can use passthrough chains.

A passthrough is a single step in a passthrough chain. The passthrough is evaluated in a sequence to incrementally build a configuration. The configuration is then passed to the target analyzer.

A configuration section for an analyzer has the following parameters:

A configuration section can include one or more passthrough sections. The maximum number of passthrough sections is 20. There are several types of passthroughs:

If multiple passthrough sections are defined in a passthrough chain, their position in the chain defines the order in which they are evaluated.

• Passthroughs listed later in the chain sequence have a higher precedence.
• Passthroughs with a higher precedence overwrite (default) and append data yielded by previous passthroughs. This is useful for cases where you need to use or modify an existing configuration.

Configure a passthrough these parameters:

The amount of data generated by a single passthrough is limited to 1MB.

Passthrough configuration examples

Raw passthrough for nodejs-scan

Define a custom analyzer configuration. In this example, customized rules are defined for the nodejs-scan scanner:

[nodejs-scan]
  description = 'custom ruleset for nodejs-scan'

  [[nodejs-scan.passthrough]]
    type  = "raw"
    value = '''
- nodejs-extensions:
  - .js

  template-extensions:
  - .new
  - .hbs
  - ''

  ignore-filenames:
- skip.js

  ignore-paths:
  - __MACOSX
  - skip_dir
  - node_modules

  ignore-extensions:
  - .hbs

  ignore-rules:
  - regex_injection_dos
  - pug_jade_template
  - express_xss

'''

File passthrough for gosec

Provide the name of the file containing a custom analyzer configuration. In this example, customized rules for the gosec scanner are contained in the file gosec-config.json:

[gosec]
  description = 'custom ruleset for gosec'

  [[gosec.passthrough]]
    type  = "file"
    value = "gosec-config.json"

Passthrough chain for semgrep

In the below example, we generate a custom configuration under the /sgrules target directory with a total timeout of 60 seconds.

Several passthrouh types generate a configuration for the target analyzer:

• Two git passthrough sections pull the head of branch refs/remotes/origin/test from the myrules Git repository, and revision 97f7686 from the sast-rules Git repository. From the sast-rules Git repository, only data from the go subdirectory is considered.
  • The sast-rules entry has a higher precedence because it appears later in the configuration.
  • If there is a filename collision between files in both repositories, files from the sast repository overwrite files from the myrules repository, as sast-rules has higher precedence.
• The raw entry creates a file named insecure.yml under /sgrules. The full path is /sgrules/insecure.yml.
• The url entry fetches a configuration made available through a URL and stores it in the /sgrules/gosec.yml file.

Afterwards, semgrep is invoked with the final configuration located under /sgrules.

[semgrep]
  description = 'semgrep custom rules configuration'
  targetdir = "/sgrules"
  timeout = 60

  [[semgrep.passthrough]]
    type  = "git"
    value = "https://gitlab.com/user/myrules.git"
    ref = "refs/remotes/origin/test"

  [[semgrep.passthrough]]
    type  = "git"
    value = "https://gitlab.com/gitlab-org/secure/gsoc-sast-vulnerability-rules/playground/sast-rules.git"
    ref = "97f7686db058e2141c0806a477c1e04835c4f395"
    subdir = "go"

  [[semgrep.passthrough]]
    type  = "raw"
    target = "insecure.yml"
    value = """
rules:
- id: "insecure"
  patterns:
  - pattern: "func insecure() {...}"
  message: |
    Insecure function insecure detected
  metadata:
    cwe: "CWE-200: Exposure of Sensitive Information to an Unauthorized Actor"
  severity: "ERROR"
  languages:
  - "go"
    """

  [[semgrep.passthrough]]
    type  = "url"
    value = "https://semgrep.dev/c/p/gosec"
    target = "gosec.yml"

Interpolation

The code snippet below shows an example configuration that uses an environment variable $GITURL to access a private repositories with a Git URL. The variable contains a username and token in the value field (for example https://user:token@url). It does not explicitly store credentials in the configuration file. To reduce the risk of leaking secrets through created paths and files, use this feature with caution.

[semgrep]
  description = 'semgrep custom rules configuration'
  targetdir = "/sgrules"
  interpolate = true

  [[semgrep.passthrough]]
    type  = "git"
    value = "$GITURL"
    ref = "refs/remotes/origin/main"

Configure the append mode for passthroughs

To append data to previous passthroughs, use the append mode for the passthrough types file, url, and raw.

Passthroughs in override mode overwrite files created when preceding passthroughs in the chain find a naming collision. If mode is set to append, a passthrough appends data to the files created by its predecessors instead of overwriting.

In the below semgrep configuration,/sgrules/insecure.yml assembles two passthroughs. The rules are:

• insecure
• secret

These rules add a search pattern to the analyzer and extends semgrep capabilities.

For passthrough chains we recommend that you enable validation. To enable validation, you can either:

• set validate to true

• set a passthrough validator to xml, json, yaml, or toml.

[semgrep]
  description = 'semgrep custom rules configuration'
  targetdir = "/sgrules"
  validate = true

  [[semgrep.passthrough]]
    type  = "raw"
    target = "insecure.yml"
    value = """
rules:
- id: "insecure"
  patterns:
  - pattern: "func insecure() {...}"
  message: |
    Insecure function insecure detected
  metadata:
    cwe: "...
  severity: "ERROR"
  languages:
  - "go"
"""

  [[semgrep.passthrough]]
    type  = "raw"
    mode  = "append"
    target = "insecure.yml"
    value = """
- id: "secret"
  patterns:
  - pattern-either:
    - pattern: "$MASK = \"...\""
  - metavariable-regex:
      metavariable: "$MASK"
      regex: "(password|pass|passwd|pwd|secret|token)"
  message: |
    Use of Hard-coded Password
    cwe: "..."
  severity: "ERROR"
  languages:
  - "go"
"""

False Positive Detection (ULTIMATE)

Introduced in GitLab 14.2.

Vulnerabilities that have been detected and are false positives will be flagged as false positives in the security dashboard.

Using CI/CD variables to pass credentials for private repositories

Some analyzers require downloading the project's dependencies in order to perform the analysis. In turn, such dependencies may live in private Git repositories and thus require credentials like username and password to download them. Depending on the analyzer, such credentials can be provided to it via custom CI/CD variables.

Using a CI/CD variable to pass username and password to a private Go repository

If your Go project depends on private modules, see Fetch modules from private projects for how to provide authentication over HTTPS.

To specify credentials via ~/.netrc provide a before_script containing the following:

gosec-sast:
  before_script:
    - |
      cat <<EOF > ~/.netrc
      machine gitlab.com
      login $CI_DEPLOY_USER
      password $CI_DEPLOY_PASSWORD
      EOF

Using a CI/CD variable to pass username and password to a private Maven repository

If your private Maven repository requires login credentials, you can use the MAVEN_CLI_OPTS CI/CD variable.

Read more on how to use private Maven repositories.

Enabling Kubesec analyzer

Introduced in GitLab 12.6.

You need to set SCAN_KUBERNETES_MANIFESTS to "true" to enable the Kubesec analyzer. In .gitlab-ci.yml, define:

include:
  - template: Security/SAST.gitlab-ci.yml

variables:
  SCAN_KUBERNETES_MANIFESTS: "true"

Pre-compilation

Most GitLab SAST analyzers directly scan your source code without compiling it first. However, for technical reasons, some analyzers can only scan compiled code.

By default, these analyzers automatically attempt to fetch dependencies and compile your code so it can be scanned. Automatic compilation can fail if:

• your project requires custom build configurations.
• you use language versions that aren't built into the analyzer.

To resolve these issues, you can skip the analyzer's compilation step and directly provide artifacts from an earlier stage in your pipeline instead. This strategy is called pre-compilation.

Pre-compilation is available for the analyzers that support the COMPILE CI/CD variable. See Analyzer settings for the current list.

To use pre-compilation:

1. Output your project's dependencies to a directory in the project's working directory, then save that directory as an artifact by setting the artifacts: paths configuration.
2. Provide the COMPILE: "false" CI/CD variable to the analyzer to disable automatic compilation.
3. Add your compilation stage as a dependency for the analyzer job.

To allow the analyzer to recognize the compiled artifacts, you must explicitly specify the path to the vendored directory. This configuration can vary per analyzer. For Maven projects, you can use MAVEN_REPO_PATH. See Analyzer settings for the complete list of available options.

The following example pre-compiles a Maven project and provides it to the SpotBugs SAST analyzer:

stages:
  - build
  - test

include:
  - template: Security/SAST.gitlab-ci.yml

build:
  image: maven:3.6-jdk-8-slim
  stage: build
  script:
    - mvn package -Dmaven.repo.local=./.m2/repository
  artifacts:
    paths:
      - .m2/
      - target/

spotbugs-sast:
  dependencies:
    - build
  variables:
    MAVEN_REPO_PATH: ./.m2/repository
    COMPILE: "false"
  artifacts:
    reports:
      sast: gl-sast-report.json

Available CI/CD variables

SAST can be configured using the variables parameter in .gitlab-ci.yml.

The following example includes the SAST template to override the SAST_GOSEC_LEVEL variable to 2. The template is evaluated before the pipeline configuration, so the last mention of the variable takes precedence.

include:
  - template: Security/SAST.gitlab-ci.yml

variables:
  SAST_GOSEC_LEVEL: 2

Logging level

Introduced in GitLab 13.1.

To control the verbosity of logs, set the SECURE_LOG_LEVEL environment variable. Messages of this logging level or higher are output.

From highest to lowest severity, the logging levels are:

• fatal
• error
• warn
• info (default)
• debug

Custom Certificate Authority

To trust a custom Certificate Authority, set the ADDITIONAL_CA_CERT_BUNDLE variable to the bundle of CA certs that you want to trust in the SAST environment. The ADDITIONAL_CA_CERT_BUNDLE value should contain the text representation of the X.509 PEM public-key certificate. For example, to configure this value in the .gitlab-ci.yml file, use the following:

variables:
  ADDITIONAL_CA_CERT_BUNDLE: |
      -----BEGIN CERTIFICATE-----
      MIIGqTCCBJGgAwIBAgIQI7AVxxVwg2kch4d56XNdDjANBgkqhkiG9w0BAQsFADCB
      ...
      jWgmPqF3vUbZE0EyScetPJquRFRKIesyJuBFMAs=
      -----END CERTIFICATE-----

The ADDITIONAL_CA_CERT_BUNDLE value can also be configured as a custom variable in the UI, either as a file, which requires the path to the certificate, or as a variable, which requires the text representation of the certificate.

Docker images

The following are Docker image-related CI/CD variables.

Vulnerability filters

Some analyzers make it possible to filter out vulnerabilities under a given threshold.

Analyzer settings

Some analyzers can be customized with CI/CD variables.

Custom CI/CD variables

Introduced in GitLab 12.5.

In addition to the aforementioned SAST configuration CI/CD variables, all custom variables are propagated to the underlying SAST analyzer images if the SAST vendored template is used.

NOTE: In GitLab 13.3 and earlier, variables whose names started with the following prefixes are not propagated to either the analyzer containers or SAST Docker container: DOCKER_, CI, GITLAB_, FF_, HOME, PWD, OLDPWD, PATH, SHLVL, HOSTNAME.

Experimental features

You can receive early access to experimental features. Experimental features might be added, removed, or promoted to regular features at any time.

Experimental features available are:

• Enable scanning of iOS and Android apps using the MobSF analyzer.

Enable experimental features

To enable experimental features, add the following to your .gitlab-ci.yml file:

include:
  - template: Security/SAST.gitlab-ci.yml

variables:
  SAST_EXPERIMENTAL_FEATURES: "true"

Reports JSON format

SAST outputs a report file in JSON format. The report file contains details of all found vulnerabilities. To download the report file, you can either:

• Download the file from the CI/CD pipelines page.
• In the pipelines tab on merge requests, set artifacts: paths to gl-sast-report.json.

For information, see Download job artifacts.

For details of the report file's schema, see SAST report file schema.

For an example SAST report file, see gl-secret-detection-report.json example.

Running SAST in an offline environment

For self-managed GitLab instances in an environment with limited, restricted, or intermittent access to external resources through the internet, some adjustments are required for the SAST job to run successfully. For more information, see Offline environments.

Requirements for offline SAST

To use SAST in an offline environment, you need:

• GitLab Runner with the docker or kubernetes executor.
• A Docker Container Registry with locally available copies of SAST analyzer images.
• Configure certificate checking of packages (optional).

GitLab Runner has a default pull_policy of always, meaning the runner tries to pull Docker images from the GitLab container registry even if a local copy is available. The GitLab Runner pull_policy can be set to if-not-present in an offline environment if you prefer using only locally available Docker images. However, we recommend keeping the pull policy setting to always if not in an offline environment, as this enables the use of updated scanners in your CI/CD pipelines.

Make GitLab SAST analyzer images available inside your Docker registry

For SAST with all supported languages and frameworks, import the following default SAST analyzer images from registry.gitlab.com into your local Docker container registry:

registry.gitlab.com/security-products/bandit:2
registry.gitlab.com/security-products/brakeman:2
registry.gitlab.com/security-products/eslint:2
registry.gitlab.com/security-products/flawfinder:2
registry.gitlab.com/security-products/gosec:3
registry.gitlab.com/security-products/kubesec:2
registry.gitlab.com/security-products/nodejs-scan:2
registry.gitlab.com/security-products/phpcs-security-audit:2
registry.gitlab.com/security-products/pmd-apex:2
registry.gitlab.com/security-products/security-code-scan:2
registry.gitlab.com/security-products/semgrep:2
registry.gitlab.com/security-products/sobelow:2
registry.gitlab.com/security-products/spotbugs:2

The process for importing Docker images into a local offline Docker registry depends on your network security policy. Please consult your IT staff to find an accepted and approved process by which external resources can be imported or temporarily accessed. These scanners are periodically updated with new definitions, and you may be able to make occasional updates on your own.

For details on saving and transporting Docker images as a file, see Docker's documentation on docker save, docker load, docker export, and docker import.

If support for Custom Certificate Authorities are needed

Support for custom certificate authorities was introduced in the following versions.

Set SAST CI/CD variables to use local SAST analyzers

Add the following configuration to your .gitlab-ci.yml file. You must replace SECURE_ANALYZERS_PREFIX to refer to your local Docker container registry:

include:
  - template: Security/SAST.gitlab-ci.yml

variables:
  SECURE_ANALYZERS_PREFIX: "localhost:5000/analyzers"

The SAST job should now use local copies of the SAST analyzers to scan your code and generate security reports without requiring internet access.

Configure certificate checking of packages

If a SAST job invokes a package manager, you must configure its certificate verification. In an offline environment, certificate verification with an external source is not possible. Either use a self-signed certificate or disable certificate verification. Refer to the package manager's documentation for instructions.

Running SAST in SELinux

By default SAST analyzers are supported in GitLab instances hosted on SELinux. Adding a before_script in an overridden SAST job may not work as runners hosted on SELinux have restricted permissions.

Troubleshooting

SAST debug logging

Increase the Secure scanner log verbosity to debug in a global CI variable to help troubleshoot SAST jobs.

variables:
  SECURE_LOG_LEVEL: "debug"

Error response from daemon: error processing tar file: docker-tar: relocation error

This error occurs when the Docker version that runs the SAST job is 19.03.0. Consider updating to Docker 19.03.1 or greater. Older versions are not affected. Read more in this issue.

Getting warning message gl-sast-report.json: no matching files

For information on this, see the general Application Security troubleshooting section.

Error: sast is used for configuration only, and its script should not be executed

For information on this, see the GitLab Secure troubleshooting section.

Limitation when using rules:exists

The SAST CI template uses the rules:exists parameter. For performance reasons, a maximum number of matches are made against the given glob pattern. If the number of matches exceeds the maximum, the rules:exists parameter returns true. Depending on the number of files in your repository, a SAST job might be triggered even if the scanner doesn't support your project. For more details about this issue, see the rules:exists documentation.

SpotBugs UTF-8 unmappable character errors

These errors occur when UTF-8 encoding isn't enabled on a SpotBugs build and there are UTF-8 characters in the source code. To fix this error, enable UTF-8 for your project's build tool.

For Gradle builds, add the following to your build.gradle file:

compileJava.options.encoding = 'UTF-8'
tasks.withType(JavaCompile) {
    options.encoding = 'UTF-8'
}

For Maven builds, add the following to your pom.xml file:

<properties>
  <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
</properties>

SpotBugs Error: Project couldn't be built

If your job is failing at the build step with the message "Project couldn't be built", it's most likely because your job is asking SpotBugs to build with a tool that isn't part of its default tools. For a list of the SpotBugs default tools, see SpotBugs' asdf dependencies.

The solution is to use pre-compilation. Pre-compilation ensures the images required by SpotBugs are available in the job's container.

Flawfinder encoding error

This occurs when Flawfinder encounters an invalid UTF-8 character. To fix this, convert all source code in your project to UTF-8 character encoding. This can be done with cvt2utf or iconv either over the entire project or per job using the before_script feature.

Semgrep slowness, unexpected results, or other errors

If Semgrep is slow, reports too many false positives or false negatives, crashes, fails, or is otherwise broken, see the Semgrep docs for troubleshooting GitLab SAST.

SAST job fails with message strconv.ParseUint: parsing "0.0": invalid syntax

Invoking Docker-in-Docker is the likely cause of this error. Docker-in-Docker is:

• Disabled by default in GitLab 13.0 and later.
• Unsupported from GitLab 13.4 and later.

Several workarounds are available. From GitLab version 13.0 and later, you must not use Docker-in-Docker.

Workaround 1: Pin analyzer versions (GitLab 12.1 and earlier)

Set the following variables for the SAST job. This pins the analyzer versions to the last known working version, allowing SAST with Docker-in-Docker to complete as it did previously:

sast:
  variables:
    SAST_DEFAULT_ANALYZERS: ""
    SAST_ANALYZER_IMAGES: "registry.gitlab.com/gitlab-org/security-products/analyzers/bandit:2.9.6, registry.gitlab.com/gitlab-org/security-products/analyzers/brakeman:2.11.0, registry.gitlab.com/gitlab-org/security-products/analyzers/eslint:2.10.0, registry.gitlab.com/gitlab-org/security-products/analyzers/flawfinder:2.11.1, registry.gitlab.com/gitlab-org/security-products/analyzers/gosec:2.14.0, registry.gitlab.com/gitlab-org/security-products/analyzers/nodejs-scan:2.11.0, registry.gitlab.com/gitlab-org/security-products/analyzers/phpcs-security-audit:2.9.1, registry.gitlab.com/gitlab-org/security-products/analyzers/pmd-apex:2.9.0, registry.gitlab.com/gitlab-org/security-products/analyzers/secrets:3.12.0, registry.gitlab.com/gitlab-org/security-products/analyzers/security-code-scan:2.13.0, registry.gitlab.com/gitlab-org/security-products/analyzers/sobelow:2.8.0, registry.gitlab.com/gitlab-org/security-products/analyzers/spotbugs:2.13.6, registry.gitlab.com/gitlab-org/security-products/analyzers/tslint:2.4.0"

Remove any analyzers you don't need from the SAST_ANALYZER_IMAGES list. Keep SAST_DEFAULT_ANALYZERS set to an empty string "".

Workaround 2: Disable Docker-in-Docker for SAST and Dependency Scanning (GitLab 12.3 and later)

Disable Docker-in-Docker for SAST. Individual <analyzer-name>-sast jobs are created for each analyzer that runs in your CI/CD pipeline.

include:
  - template: SAST.gitlab-ci.yml

variables:
  SAST_DISABLE_DIND: "true"

Workaround 3: Upgrade to GitLab 13.x and use the defaults

From GitLab 13.0, SAST defaults to not using Docker-in-Docker. In GitLab 13.4 and later, SAST using Docker-in-Docker is no longer supported. If you have this problem on GitLab 13.x and later, you have customized your SAST job to use Docker-in-Docker. To resolve this, comment out any customizations you've made to your SAST CI job definition and follow the documentation to reconfigure, using the new and improved job definition default values.

include:
  - template: Security/SAST.gitlab-ci.yml