Skip to content

Files

Latest commit

 

History

History
740 lines (526 loc) · 35.8 KB

README.md

File metadata and controls

740 lines (526 loc) · 35.8 KB

RapiDAST

GitHub Actions Workflow Status GitHub License

RapiDAST (Rapid DAST) is an open-source security testing tool that automates DAST (Dynamic Application Security Testing) and streamlines the integration of security testing into development workflows. It is designed to help Developers and/or QA engineers rapidly and effectively identify low-hanging security vulnerabilities in your applications, ideally in CI/CD pipelines. RapiDAST is for organizations implementing DevSecOps with a shift-left approach.

RapiDAST provides:

  • Automated HTTP/API security scanning using ZAP
  • Kubernetes operator scanning using OOBTKUBE
  • Automated vulnerability scanning using Nessus (requires a Nessus instance)
  • Command-line execution with yaml configuration, suitable for integration in CI/CD pipelines
  • Ability to run automated DAST scanning with pre-built or custom container images
  • HTML, JSON and XML report generation
  • Integration with Google Cloud Storage and OWASP DefectDojo

RapiDAST is for testing purposes, and should not be used on production systems.

Quickstart

Quickly setup RapiDAST to scan a target application. See Workflow for more information.

  1. Create a minimal config file for the target application, see Configuration section for details
  2. Run RapiDAST with the config file, either in a container or from source code

OS Support

Linux and MacOS are both supported, however running RapiDAST in a container is currently only supported on Linux. See MacOS Configuration section for more details.

Run in container (Linux only)

Run the pre-built rapidast container image, which includes scanners like ZAP. Not compatible with config files using general.container.type set to podman.

Prerequisites

  • docker / podman (>= v3.0.1)

Run

$ podman run -v ./config.yaml:/opt/rapidast/config/config.yaml:Z quay.io/redhatproductsecurity/rapidast:latest ./rapidast.py

Note

  • Sample config is very minimal and has no Authentication enabled
  • The :Z option is only necessary on RHEL/CentOS/Fedora systems with SELinux enabled
  • To retrieve scan results, add a volume mount like -v ./results/:/opt/rapidast/results/:Z. The permissions of the ./results/ directory may need to be modified first with a command like chmod o+w ./results/ to be writeable by the rapidast user in the container.

Run from source

Install dependencies and run RapiDAST directly on a host machine. Unless using the config setting of general.container.type: podman, scanners like ZAP are expected to be installed on the host system.

Prerequisites

  • python >= 3.6.8 (3.7 for MacOS/Darwin)
  • podman >= 3.0.1
    • required when you want to run scanners from their container images, rather than installing them to your host.
  • See requirements.txt for a list of required python libraries

Setup

Clone the repository.

$ git clone https://github.com/RedHatProductSecurity/rapidast.git
$ cd rapidast

Create a virtual environment.

$ python3 -m venv venv
$ source venv/bin/activate

Install the project requirements.

(venv) $ pip install -U pip
(venv) $ pip install -r requirements.txt

Run

Run RapiDAST script:

$ ./rapidast.py --config <path/to/config.yml>

Note

  • Example minimum config expects scanners like ZAP to be available on the host, and will fail if not found. See Execution Environments section for more info
  • Results will be written to the ./results/ directory

Workflow

This section summarize the basic workflow as follows:

  1. Create a configuration file for testing the application. See the configuration section below for more information.
    • Optionally, an environment file may be added, e.g., to separate the secrets from the configuration file.
  2. Run RapiDAST and get the results.
    • First run with passive scanning only, which can save time at the initial scanning phase. There are various situations that can cause an issue, not only from scanning set up but also from your application or test environment. Active scanning takes a long time in general.
    • Once passive Scanning has run successfully, run another scan with active scanning enabled in the configuration file.

Configuration

The configuration file is presented as YAML, and contains several main entries:

  • config : contains configVersion which tells RapiDAST how to consume the config file
  • application : contains data relative to the application being scanned : name, etc.
  • general : contains data that will be used by all the scanners, such as proxy configuration, etc.
    • Each scanner can override an entry from general by creating an entry with the same name
  • scanners : list of scanners, and their configuration

See templates in the config directory for examples and ideas.

  • config-template-zap-tiny.yaml : describes a bare minimum configuration, without authentication options.
  • config-template-zap-simple.yaml : describes a generic/minimal use of the ZAP scanner (i.e.: the minimum set of option to get a ZAP scan from RapiDAST)
  • config-template-zap-mac.yaml : describes a minimal use of the ZAP scanner on a Apple Mac environment
  • config-template-zap-long.yaml : describes a more extensive use of ZAP (all configuration options are presented)
  • config-template-multi-scan.yaml : describes how to combine multiple scanners in a single configuration
  • config-template-generic-scan.yaml : describes the use of the generic scanner

Basic Example

Example bare minimum config file, without any Authentication options, and passive scanning only:

config:
  configVersion: 5

application:
  shortName: "example-1.0"
  url: "https://example.com" # root URL of the application

scanners:
  zap:
    apiScan:
      apis:
        apiUrl: "https://example.com/api/v1/swagger.json" # URL to application openAPI spec

Authentication

Authentication is configured in the general entry, as it can be applied to multiple scanning options. Currently, Authentication is applied to ZAP scanning only. In the long term it may be applied to other scanning configurations.

Supported options:

  • No authentication: The scanners will communicate anonymously with the application

  • OAuth2 using a Refresh Token: This method describes required parameters needed to retrieve an access token, using a refresh token as a secret.

    • authentication type : oauth2_rtoken
    • parameters :
      • token_endpoint: the URL to which send the refresh token
      • client_id : the client ID
      • rtoken_var_name: for practical reasons, the refresh token is provided using environment variables. This entry describes the name of the variable containing the secret refresh token
      • preauth: Pre-generate a token and force ZAP to use it throughout the session (the session token will not be refreshed after it's expired). Default: False. This is only useful for scans sufficiently short that it will be finished before the token expires
  • HTTP Basic: This method describes the HTTP Basic Authorization Header. The username and password must be provided in plaintext and will be encoded by the scanners

    • authentication type: http_basic
    • parameters:
      • username
      • password
  • HTTP Header: This method describes the HTTP generic header. The name and value must be provided in plaintext.

    • authentication type: http_header
    • parameters:
      • name: the header name added to every request. By default is Authorization
      • value or value_from_var (the environment variable with the secret)
  • Cookie Authentication: This method describes authentication via Cookie header. The cookie name and value must be provided in plaintext.

    • authentication type: cookie
    • parameters:
      • name
      • value
  • Browser authentication method This method uses firefox in the background to load a login page and fill in username/password, and will retrieve and set the session cookies accordingly.

    • authentication type: browser
    • parameters:
      • username
      • password
      • loginPageUrl: the URL to the login page (either the full URL, or relative to the application.url value)
      • verifyUrl: a URL that "proves" the user is authenticated (either the full URL, or relative to the application.url value). This URL must return a success if the user is correctly authenticated, and an error otherwise.

MacOS

RapiDAST supports executing scanners like ZAP on the MacOS host directly only.

To run RapiDAST on MacOS(See the Configuration section below for more details on configuration):

  • Set general.container.type: "none" or scanners.zap.container.type: "none" in the configuration.
  • Configure scanners.zap.container.parameters.executable to the installation path of the zap.sh command, because it is not available in the PATH. Usually, its path is /Applications/ZAP.app/Contents/Java/zap.sh on MacOS.

Example:

scanners:
  zap:
    container:
      type: none
      parameters:
        executable: "/Applications/ZAP.app/Contents/Java/zap.sh"

Advanced configuration

You may not want to directly have configuration values inside the configuration. Typically: either the entry is a secret (such as a password), but the configuration needs to be public, or the entry needs to be dynamically generated (e.g.: a cookie, a uniquely generated URL, etc.) at the time of running RapiDAST, and it's an inconvenient to always having to modify the configuration file for each run.

To avoid this, RapiDAST proposes 2 ways to provide a value for a given configuration entry. For example, to provide a value for the entry general.authentication.parameters.rtoken, you can either (in order of priority):

  • Create an entry in the configuration file (this is the usual method)
  • Create an entry in the configuration file pointing to the environment variable that actually contains the data, by appending _from_var to the entry name: general.authentication.parameters.rtoken_from_var=RTOKEN (in this example, the token value is provided by the $RTOKEN environment variable)

Running several instance of a scanner

It is possible to run a scanner several times with different configurations. This is done by adding a different identifier to each scan, by appending _<id> to the scanner name.

For example :

scanners:
  zap_unauthenticated:
    apiScan:
      apis:
        apiUrl: "https://example.com/api/openapi.json"

  zap_authenticated:
    authentication:
      type: "http_basic"
      parameters:
        username: "user"
        password: "mypassw0rd"
    apiScan:
      apis:
        apiUrl: "https://example.com/api/openapi.json"

In the example above, the ZAP scanner will first run without authentication, and then rerun again with a basic HTTP authentication. The results will be stored in their respective names (i.e.: zap_unauthenticated and zap_authenticated in the example above).

Exporting data to external services

Exporting to Google Cloud Storage

This simply stores the data as a compressed tarball in a Google Cloud Storage bucket.

config:
  # Defect dojo configuration
  googleCloudStorage:
    keyFile: "/path/to/GCS/key"                           # optional: path to the GCS key file (alternatively: use GOOGLE_APPLICATION_CREDENTIALS)
    bucketName: "<name-of-GCS-bucket-to-export-to>"       # Mandatory
    directory: "<override-of-default-directory>"          # Optional directory where the credentials have write access, defaults to `RapiDAST-<product>`

Once this is set, scan results will be exported to the bucket automatically. The tarball file will include:

  1. metadata.json - the file that contains scan_type, uuid and import_data(could be changed later. Currently this comes from the previous DefectDojo integration feature)
  2. scans - the directory that contains scan results

Exporting to DefectDojo

RapiDAST supports integration with OWASP DefectDojo which is an open source vulnerability management tool.

Preamble: creating DefectDojo user

RapiDAST needs to be able to authenticate to your DefectDojo instance. However, ideally, it should have the minimum set of permissions, such that it will not be allowed to modify products other than the one(s) it is supposed to.

In order to do that:

  • create a user without any global role
  • add that user as a "writer" for the product(s) it is supposed to scan

Then the product, as well as an engagement for that product, must be created in your DefectDojo instance. It would not be advised to give the RapiDAST user an "admin" role and simply set auto_create_context to True, as it would be both insecure and accident prone (a typo in the product name would let RapiDAST create a new product)

Exporting to Defect Dojo

RapiDAST will send the results directly to a DefectDojo service. This is a typical configuration:

config:
  # Defect dojo configuration
  defectDojo:
    url: "https://mydefectdojo.example.com/"
    ssl: [True | False | "/path/to/CA"]
    authorization:
      username: "rapidast_productname"
      password: "password"
      # alternatively, a `token` entry can be set in place of username/password

The ssl parameter is provided as the Python Requests module's verify parameter. It can be either:

  • True: SSL verification is mandatory, against the default CA bundle
  • False: SSL verification is not mandatory (but prints a warning if it fails)
  • /path/to/CA: a bundle of CAs to verify from Alternatively, the REQUESTS_CA_BUNDLE environment variable can be used to select a CA bundle file. If nothing is provided, the default value will be True

You can either authenticate using a username/password combination, or a token (make sure it is not expired). In either case, you can use the _from_var method described in the previous chapter to avoid hardcoding the value in the configuration.

Configuration of exported data

The data exported follows the Defectdojo methodology of "Product → Engagement → Test" : a test, such as a ZAP scan, belongs to an engagement for a product. Its configuration is made under the scanners.<scanner>.defectDojoExport.parameters configuration entries. As a baseline, parameters from the Defectdojo import-scan and reimport-scan are accepted.

For each scan, the logic applied is the following, in order:

  • If a test ID is provided (parameter test), this scan will replace the previous one (a "reimport" in Defectdojo)
  • If an engagement ID is provided (parameter engagement), this scan will be added as a new test in that existing engagement
  • If an engagement and a product are given by name (engagement_name and product_name parameters), this scan will be added for that given engagement for the given product

In each defectDojoExport.parameters, some defaults parameters are applied:

  • product_name, in order (the first non empty value found):
    • application.productName
    • application.shortName (this name should not contain non-printable characters, such as spaces)
  • engagement_name defaults to RapiDAST-<product name>-<date>
  • scan_type : filled by the scanner
  • active: True
  • verified: False

As a reminder: values from general are applied to each scanner.

Here is an example:

scanners:
  zap:
    defectDojoExport:
      parameters:
        product_name: "My Product"
        engagement_name: "RapiDAST" # or engagement: <engagement_id>
        #test: <test_id>

See https://documentation.defectdojo.com/integrations/importing/#api for more information.

Execution

Once you have created a configuration file, you can run a scan with it.

$ ./rapidast.py --config "<your-config.yaml>"

There are more options.

usage: rapidast.py [-h] [--log-level {debug,info,warning,error,critical}]
                   [--config CONFIG_FILE] [--no-cleanup]

Runs various DAST scanners against a defined target, as configured by a
configuration file.

options:
  -h, --help            show this help message and exit
  --log-level {debug,info,warning,error,critical}
                        Level of verbosity
  --config CONFIG_FILE  Path to YAML config file
  --no-cleanup          Scanners to not cleanup their environment. (might be
                        useful for debugging purposes).

Choosing the execution environment

Set general.container.type to select an environment (default: none)

  • none (default):

    • Run a RapiDAST scan with scanners that are installed on the same host OR run RapiDAST in a container (scanners are to be installed in the same container image)
    • Warning: without a container layer, RapiDAST may have to modify the host's file system, such as the tools configuration to fit its needs. For example: the ZAP plugin has to copy the policy file used in ZAP's user config directory (~/.ZAP)
  • podman:

    • Run scanners as separate containers using podman
    • RapiDAST must not run inside a container
    • Select the image to load from scanner.<name>.container.image (sensible default are provided for each scanner)

It is also possible to set the container type for each scanner differently by setting scanners.<name>.container.type under a certain scanner configuration. Then the scanner will run from its image, regardless of the general.container.type value.

Build a RapiDAST image

If you want to build your own RapiDAST image, run the following command.

$ podman build . -f containerize/Containerfile -t <image-tag>

Disclaimer: This tool is not intended to be run as a long-running service. Instead, it is designed to be run for a short period of time while a scan is being invoked and executed in a separate test environment. If this tool is used solely for the scanning purposes, vulnerabilities that may be indicated to exist in the image will not have a chance to be exploited. The user assumes all risks and liability associated with its use.

Running on Kubernetes or OpenShift

Helm chart is provided to help with running RapiDAST on Kubernetes or OpenShift.

See helm/README.md

Scanners

ZAP

ZAP (Zed Attack Proxy) is an open-source DAST tool. It can be used for scanning web applications and API.

See https://www.zaproxy.org/ for more information.

Methodology

ZAP needs to be pointed to a list of endpoints to the tested application. Those can be:

  • A regular HTML page
  • A REST endpoint
  • A GraphQL interface

The GraphQL interface can be provided to RapiDAST via the graphql configuration entry. It requires the URL of the GraphQL interface and the GraphQL schema(if available), in order to be scanned. Additional options are available. See the config-template-zap-long.yaml configuration template file for a list of options.

The other endpoints can be provided via several methods, discussed in the chapters below.

an OpenAPI schema

This is the prefered method, to be used whenever possible. RapiDAST accepts OpenAPI v2(formerly known as Swagger) and v3 schemas. These schemas will describe a list of endpoints, and for each of them, a list of parameters accepted by the application.

Build the endpoint list using a spider/crawler

In this method, RapiDAST is given a Web entrypoint. The crawler will download that page, extract a list of URLs and recursively crawl all of them. The entire list of URLs found is then provided to the scanner.

There are two crawlers available:

  • Basic spider: the list of URLs will be searched in the HTML tags (e.g.: <a>, <img>, etc.)
  • Ajax spider: this crawler will run a real browser (by default: firefox headless), allowing the dynamic execution of Javascripts from each page found. This method will find URLs generated dynamically.

See the spider and spiderAjax configuration entries in the config-template-zap-long.yaml configuration template file for a list of options available.

A list of endpoints

A file containing a list of URLs corresponding to endpoints and their parameters.

Example of file:

https://example.com/api/v3/groupA/functionA?parameter1=abc&parameter2=123
https://example.com/api/v3/groupB/functionB?parameter1=def&parameter2=456

Only GET requests will be scanned.

ZAP scanner specific options

Below are some configuration options that are worth noting, when running a RapiDAST scan with the ZAP scanner.

  • (*.container.type: podman only) Inject the ZAP container in an existing Pod:

It is possible to gather both RapiDAST and the tested application into the same podman Pod and run a scan against the application. This might help CI/CD automation & clean-up. In order to do that, the user must create the Pod prior to running RapiDAST, and indicate its name in the RapiDAST configuration: scanners.zap.container.parameters.podName. However, it is currently necessary to map the host user to UID 1000 / GID 1000 manually during the creation of the Pod using the --userns=keep-id:uid=1000,gid=1000 option Example: podman pod create --userns=keep-id:uid=1000,gid=1000 myApp_Pod

  • (when running scans on the desktop with the *.container.type: none configuration only) Enable ZAP's Graphical UI:

This is useful for debugging. Set scanners.zap.miscOptions.enableUI: True (default: False). Then, the ZAP desktop will run with GUI on your host and show the progress of scanning.

  • Enable add-on updates:

Set scanners.zap.miscOptions.updateAddons: True (default: False). ZAP will first update its addons and then run the scan.

  • Install additional addons:

Set scanners.zap.miscOptions.additionalAddons: "comma,separated,list,of,addons" (default: []). Prior to running a scan, ZAP will install a given list of addons. The list can be provided either as a YAML list, or a string of the addons, separated by a comma.

  • Force maximum heap size for the JVM:

Set scanners.zap.miscOptions.memMaxHeap (default: ¼ of the RAM), similarly to Java's -Xmx option.

Example:

scanners:
    zap:
        container:
            parameters:
                podName: "myApp_Pod"
        miscOptions:
            enableUI: True
            updateAddons: False
            memMaxHeap: "6144m"
  • To use ZAP's '-config' option:

Set scanners.zap.miscOptions.overrideConfigs with the same value as you would run with ZAP's '-config' option. It allows RapiDAST to run additional '-config' options when it invokes the ZAP cli command. This can be useful to set a value for Path parameters of the OpenAPI specification. The following example will allow RapiDAST to send the 'default' value to the {namespace} parameter in your OpenAPI file.

Example:

scanners:
    zap:
    	overrideConfigs:
        	- formhandler.fields.field(0).fieldId=namespace
        	- formhandler.fields.field(0).value=default

Nessus

Nessus is a vulnerability scanner developed by Tenable, Inc. It helps organizations identify and address security vulnerabilities across various systems, devices, and applications.

The following is an example to launch a scan:

scanners:
  nessus:
    server:
      url: https://nessus-example.com/ # URL of Nessus instance
      username: foo # OR username_from_var: NESSUS_USER
      password: bar # OR password_from_var: NESSUS_PASSWORD
    scan:
      name: test-scan # name of new scan to create
      folder: test-folder # name of folder in to contain scan
      policy: "py-test" # policy used for scan
      # timeout: 600 # timeout in seconds to complete scan
      targets:
      - 127.0.0.1

Generic scanner

RapiDAST can run other scanning tools as well as ZAP. It is possible to request RapiDAST to run a command and process stdout results, using the generic plugin.

The following is an example to run a command or a tool in the host where a RapiDAST scan runs:

scanners:
  generic:
    results: "*stdout"

    # this config is used when container.type is not 'podman'
    toolDir: scanners/generic/tools
    inline: "echo 'any scan'"

(an experimental feature) The following example is to scan a Kubernetes Operator's controller code for a command injection attack:

scanners:
  generic:
    results: "*stdout"

    # this config is used when container.type is not 'podman'
    # toolDir: scanners/generic/tools
    inline: "python3 oobtkube.py -d 300 -p <port> -i <ipaddr> -f <cr_example>.yaml"

The following is another example to run a Trivy scan using the container image:

scanners:
  generic:
    results: "*stdout"

    container:
      type: "podman"
      parameters:
        image: "docker.io/aquasec/trivy"
        command: "image docker.io/aquasec/trivy"

The results entry works as follow:

  • if it is missing or *stdout, the output of the command will be chosen and stored as stdout-report.txt in the result directory
  • if it is a directory, it will be recursively copied into the result directory
  • if it is a file, it will be copied into the result directory

Notes:

  • command can be either a list of string, or a single string which will be split using shlex.split() - when using *.container.type: podman, the results (if different from stdout) must be present on the host after podman has run, which likely means you will need to use the container.parameters.volumes entry to share the results between the container and the host.
  • See config/config-template-generic-scan.yaml for additional options.

Troubleshooting

Hitting docker.io rate limits

If you are unable to pull/update an image from docker.io due to rate-limit errors, authenticate to your Docker Hub account.

"Error getting access token" using OAuth2

Possible pitfalls :

  • Make sure that the parameters are correct (client_id, token_endpoint, rtoken_var_name) and that the refresh token is provided (via environment variable), and is valid
  • Make sure you do not have an environment variable in your current environment that overrides what is set in the envFile

Issues with the ZAP scanner

The best way to start is to look at the ZAP logs, which are stored in ~/.ZAP/zap.log (within the container where ZAP was running)

Example with podman, considering that the container was not wiped (either --no-cleanup, or the container failed):

[rapidast-ng]$ podman container list --all
969d721cc5a8  docker.io/owasp/zap2docker-stable:latest  /zap/zap.sh -conf...  2 days ago   Exited (1) 2 days ago (unhealthy)              rapidast_zap_vapi_JxgLjx
[rapidast-ng]$ podman unshare
bash-5.2# podman mount rapidast_zap_vapi_JxgLjx
/home/cedric/.local/share/containers/storage/overlay/a5450de782fb7264ff4446d96632e6512e3ff2275fd05329af7ea04106394b42/merged
bash-5.2# cd /home/cedric/.local/share/containers/storage/overlay/a5450de782fb7264ff4446d96632e6512e3ff2275fd05329af7ea04106394b42/merged
bash-5.2# tail home/zap/.ZAP/zap.log

org.zaproxy.zap.extension.openapi.converter.swagger.SwaggerException: Failed to parse swagger defn null
2023-02-17 22:42:55,922 [main ] INFO  CommandLine - Job openapi added 1 URLs
2023-02-17 22:42:55,922 [main ] INFO  CommandLine - Job openapi finished
2023-02-17 22:42:55,923 [main ] INFO  CommandLine - Automation plan failures:
2023-02-17 22:42:55,923 [main ] INFO  CommandLine -     Job openapi target: https://vapi.example.com/api/vapi/v1 error: Failed to parse OpenAPI definition.

org.zaproxy.zap.extension.openapi.converter.swagger.SwaggerException: Failed to parse swagger defn null
2023-02-17 22:42:55,924 [main ] INFO  Control - Automation Framework setting exit status to due to plan errors
2023-02-17 22:43:01,073 [main ] INFO  CommandLineBootstrap - OWASP ZAP 2.12.0 terminated.

ZAP's plugins are missing from the host installation

This happens only when using the host's ZAP (with the *.container.type: none option).

If you see a message such as Missing mandatory plugins. Fixing, or ZAP fails with an error containing the string The mandatory add-on was not found:, this is because ZAP deleted the application's plugin. See zaproxy/zaproxy#7703 for additional information. RapiDAST works around this bug, but with little inconvenients (slower because it has to fix itself and download all the plugins)

  • Verify that the host installation directory is missing its plugins. e.g., in a MacOS installation, /Applications/ZAP.app/Contents/Java/plugin/ will be mostly empty. In particular, no callhome*.zap and network*.zap file are present.
  • Reinstall ZAP, but DO NOT RUN IT, as it would delete the plugins. Verify that the directory contains many plugins.
  • chown the installation files to root, so that when running ZAP, the application running as the user does not have sufficient permission to delete its own plugins

ZAP crashing with java.lang.OutOfMemoryError: Java heap space

ZAP allows the JVM heap to grow up to a quarter of the RAM. The value can be increased using the scanners.zap.miscOptions.memMaxHeap configuration entry

2023-09-04 08:44:37,782 [main ] INFO  CommandLine - Job openapi started
2023-09-04 08:44:46,985 [main ] INFO  CommandLineBootstrap - OWASP ZAP 2.13.0 terminated.
2023-09-04 08:44:46,985 [main ] ERROR UncaughtExceptionLogger - Exception in thread "main"
java.lang.OutOfMemoryError: Java heap space
        at java.lang.AbstractStringBuilder.<init>(AbstractStringBuilder.java:86) ~[?:?]
        at java.lang.StringBuilder.<init>(StringBuilder.java:116) ~[?:?]
        at com.fasterxml.jackson.core.util.TextBuffer.contentsAsString(TextBuffer.java:487) ~[?:?]
        at com.fasterxml.jackson.core.io.SegmentedStringWriter.getAndClear(SegmentedStringWriter.java:99) ~[?:?]
        at com.fasterxml.jackson.databind.ObjectWriter.writeValueAsString(ObjectWriter.java:1141) ~[?:?]
        at io.swagger.v3.core.util.Json.pretty(Json.java:24) ~[?:?]
        at org.zaproxy.zap.extension.openapi.ExtensionOpenApi.importOpenApiDefinitionV2(ExtensionOpenApi.java:371) ~[?:?]
        at org.zaproxy.zap.extension.openapi.automation.OpenApiJob.runJob(OpenApiJob.java:123) ~[?:?]
        at org.zaproxy.addon.automation.ExtensionAutomation.runPlan(ExtensionAutomation.java:366) ~[?:?]
        at org.zaproxy.addon.automation.ExtensionAutomation.runAutomationFile(ExtensionAutomation.java:507) ~[?:?]
        at org.zaproxy.addon.automation.ExtensionAutomation.execute(ExtensionAutomation.java:621) ~[?:?]
        at org.parosproxy.paros.extension.ExtensionLoader.runCommandLine(ExtensionLoader.java:553) ~[zap-2.13.0.jar:2.13.0]
        at org.parosproxy.paros.control.Control.runCommandLine(Control.java:426) ~[zap-2.13.0.jar:2.13.0]
        at org.zaproxy.zap.CommandLineBootstrap.start(CommandLineBootstrap.java:91) ~[zap-2.13.0.jar:2.13.0]
        at org.zaproxy.zap.ZAP.main(ZAP.java:94) ~[zap-2.13.0.jar:2.13.0]

ZAP crashed while parsing the OpenAPI due to its size

2024-02-29 19:35:24,526 [main ] INFO  CommandLine - Job openapi started
2024-02-29 19:35:25,576 [main ] WARN  DeserializationUtils - Error snake-parsing yaml content
io.swagger.v3.parser.util.DeserializationUtils$SnakeException: Exception safe-checking yaml content  (maxDepth 2000, maxYamlAliasesForCollections 2147483647)
    at io.swagger.v3.parser.util.DeserializationUtils$CustomSnakeYamlConstructor.getSingleData(DeserializationUtils.java:483) ~[openapi-beta-37.zap:?]
    at org.yaml.snakeyaml.Yaml.loadFromReader(Yaml.java:493) ~[openapi-beta-37.zap:?]
.........

2024-02-29 19:35:25,639 [main ] ERROR DeserializationUtils - Error parsing content
com.fasterxml.jackson.dataformat.yaml.JacksonYAMLParseException: The incoming YAML document exceeds the limit: 3145728 code points.
 at [Source: (StringReader); line: 49813, column: 50]
..........

2024-02-29 19:35:25,702 [main ] WARN  OpenAPIV3Parser - Exception while parsing:
com.fasterxml.jackson.dataformat.yaml.JacksonYAMLParseException: The incoming YAML document exceeds the limit: 3145728 code points.
 at [Source: (StringReader); line: 49813, column: 50]

Solutions:

  • If you are using a Swagger v2 definition, try converting it to v3 (OpenAPI)
  • Set a maxYamlCodePoints Java proprety with a big value, which can be passed using environment variables (via the config.environ.envFile config entry): _JAVA_OPTIONS=-DmaxYamlCodePoints=99999999

ZAP's Ajax Spider failing

Insufficient Resources

Zap's Ajax Spider makes use of a lot of resources, in particular:

  • Shared Memory (/dev/shm)
  • processes

If you see evidence of Firefox crashing, either via in the zap.log files stored in session.tar.gz file (see below for examples of such evidence), or logged by an external crash report (such as abrtd for example).

zap.log hints for Firefox crashing:

2024-07-04 11:21:32,061 [ZAP-AjaxSpiderAuto] WARN  SpiderThread - Failed to start browser firefox-headless
com.google.inject.ProvisionException: Unable to provision, see the following errors:

1) [Guice/ErrorInCustomProvider]: SessionNotCreatedException: Could not start a new session. Response code 500. Message: Failed to decode response from marionette

Or the following:

2024-07-04 12:23:28,027 [ZAP-AjaxSpiderAuto] ERROR UncaughtExceptionLogger - Exception in thread "ZAP-AjaxSpiderAuto"
java.lang.OutOfMemoryError: unable to create native thread: possibly out of memory or process/resource limits reached

This issue may also be apparent outside of the spider, in particular, the following error being printed on the RapiDAST output is likely an evidence that the maximum number of concurrent thread is currently reached:

Failed to start thread "Unknown thread" - pthread_create failed (EAGAIN) for attributes: stacksize: 1024k, guardsize: 0k, detached.

Solutions:

  • Selenium, used to control Firefox, uses shared memory (/dev/shm/). When using the RapiDAST image or the ZAP image, the user needs to make sure that sufficient space is available in /dev/shm/ (in podman, by default, its size is 64MB). A size of 2G is the recommended value by the Selenium community. In podman for example, the option would be --shm-size=2g.
  • Zap and Firefox can create a huge numbers of threads. Some container engines will default to 2048 concurrent pids, which is not sufficient for the Ajax Spider. Whenever possible, RapiDAST will check if that limit was reached, after the scan is finished, and prints a warning if this happened. In podman, increasing the maximum number of concurrent pids is done via the --pids-limit=-1 option to prevent any limits.

Podman errors

subuid/subgid are not enabled

If you see one of those errors:

Error: copying system image from manifest list: writing blob: adding layer with blob "sha256:82aabceedc2fbf89030cbb4ff98215b70d9ae35c780ade6c784d9b447b1109ed": processing tar file(potentially insufficient UIDs or GIDs available in user namespace (requested 0:42 for /etc/gshadow): Check /etc/subuid and /etc/subgid if configured locally and run "podman system migrate": lchown /etc/gshadow: invalid argument): exit status 1

-or-

Error: parsing id map value "-1000": strconv.ParseUint: parsing "-1000": invalid syntax

Podman, in rootless mode (running as a regular user), needs subuid/subgit to be enabled: rootless mode

Caveats

  • Currently, RapiDAST does not clean up the temporary data when there is an error. The data may include:
    • a /tmp/rapidast_*/ directory
    • a podman container which name starts with rapidast_

This is to help with debugging the error. Once confirmed, it is necessary to manually remove them.

Support

If you encounter any issues or have questions, please open an issue on GitHub.

Contributing

Contribution to the project is more than welcome.

See CONTRIBUTING.md