Skip to content

Latest commit

 

History

History
123 lines (103 loc) · 7.46 KB

attack-tree.md

File metadata and controls

123 lines (103 loc) · 7.46 KB

Attack Tree Analysis for apache/airflow

Objective: Gain Unauthorized Access to Sensitive Data/Resources or Disrupt Critical Business Processes Orchestrated by Airflow

Attack Tree Visualization

                                     +-----------------------------------------------------+
                                     |  Gain Unauthorized Access/Disrupt Critical Processes |
                                     +-----------------------------------------------------+
                                                        |
          +-----------------------------------------------------------------------------------------------------------------+
          |                                                |                                                                |
+-------------------------+                 +-----------------------------+                                +-------------------------+
|  Exploit Airflow Core   |                 |  Exploit Airflow Providers  |                                |  Exploit Airflow Config  |
|       Components        |                 |        (e.g., AWS, GCP)     |                                |        & Deployment      |
+-------------------------+                 +-----------------------------+                                +-------------------------+
          |                                                |                                                                |
+---------+---------+---------+          +---------+---------+---------+---------+                      +---------+---------+---------+
|  Web UI | Scheduler | Executor |          |  Weak   |  Vuln.  |  Mis-   |  Leaked |                      |  Weak   |  Exposed|  Default|
|  Vulns |  Vulns  |  Vulns  |          |  Auth   |  Code   |  Config |  Creds  |                      |  Creds  |  API    |  Creds  |
+---------+---------+---------+          +---------+---------+---------+---------+                      +---------+---------+---------+
    |         |         |                      |         |         |         |                            |         |         |
+---+---+ +---+---+ +---+---+              +---+---+ +---+---+ +---+---+ +---+---+                    +---+---+ +---+---+ +---+---+
|XSS/  |[HIGH-RISK]|DoS/  |[HIGH-RISK]|Code   |[CRITICAL]     |AWS    |[HIGH-RISK]|AWS    | |AWS    |[HIGH-RISK]|AWS    |[HIGH-RISK]|DB     |[HIGH-RISK]|Web UI |[HIGH-RISK]|Admin  |[CRITICAL]
|CSRF  |          |Race  |          |Exec.  |      |S3     |          |Lambda | |EC2    | |Secrets|                    |Creds  |          |API    |          |PW     |
|in UI |          |Cond. |          |in     |      |Bucket |          |Vuln.  |[HIGH-RISK]|Mis-   |          |Manager|                    |       |          |Key    |          |       |
|       |          |      |          |Sched. |      |Access |          |       |          |config |          |       |[HIGH-RISK]          |       |          |       |          |       |
+---+---+ +---+---+ +---+---+              +---+---+ +---+---+ +---+---+ +---+---+                    +---+---+ +---+---+ +---+---+

Attack Tree Path: Exploit Airflow Core Components

  • Web UI Vulnerabilities:

    • XSS/CSRF in UI [HIGH-RISK]:
      • Description: Cross-Site Scripting (XSS) allows attackers to inject malicious scripts into the web interface, while Cross-Site Request Forgery (CSRF) allows attackers to trick users into performing actions they didn't intend.
      • Likelihood: Medium
      • Impact: High
      • Effort: Low to Medium
      • Skill Level: Intermediate
      • Detection Difficulty: Medium

  • Scheduler Vulnerabilities:

    • DoS/Race Conditions [HIGH-RISK]:
      • Description: Denial-of-Service (DoS) attacks overwhelm the scheduler, preventing tasks from being scheduled. Race conditions can lead to unexpected behavior or data corruption due to concurrent access.
      • Likelihood: Medium
      • Impact: Medium to High
      • Effort: Low to Medium
      • Skill Level: Intermediate
      • Detection Difficulty: Easy to Medium

  • Executor Vulnerabilities:

    • Code Execution in Scheduler/Executor [CRITICAL]:
      • Description: Attackers inject malicious code into DAG definitions or tasks, which is then executed by the scheduler or executor without proper sandboxing. This is the most dangerous vulnerability.
      • Likelihood: Medium
      • Impact: Very High
      • Effort: Medium to High
      • Skill Level: Advanced
      • Detection Difficulty: Medium to Hard

Attack Tree Path: Exploit Airflow Providers

  • Weak Authentication (AWS S3 Bucket Access) [HIGH-RISK]:

    • Description: Airflow's connection to an AWS S3 bucket uses weak or compromised credentials, allowing unauthorized access to data.
    • Likelihood: Medium
    • Impact: High
    • Effort: Low
    • Skill Level: Novice
    • Detection Difficulty: Medium

  • Vulnerable Code (AWS Lambda Vulnerability) [HIGH-RISK]:

    • Description: An Airflow task uses a vulnerable AWS Lambda function, which the attacker exploits to gain access to other AWS resources.
    • Likelihood: Medium
    • Impact: High
    • Effort: Medium
    • Skill Level: Intermediate to Advanced
    • Detection Difficulty: Medium

  • Misconfiguration (AWS EC2 Misconfiguration) [HIGH-RISK]:

    • Description: An Airflow task interacts with a misconfigured AWS EC2 instance (e.g., open security groups), allowing unauthorized access.
    • Likelihood: Medium
    • Impact: High
    • Effort: Low to Medium
    • Skill Level: Intermediate
    • Detection Difficulty: Medium

  • Leaked Credentials (AWS Secrets Manager) [HIGH-RISK]:

    • Description: Airflow's access to AWS Secrets Manager is compromised (e.g., leaked API key), allowing retrieval of sensitive credentials.
    • Likelihood: Low to Medium
    • Impact: Very High
    • Effort: Low
    • Skill Level: Novice
    • Detection Difficulty: Hard

Attack Tree Path: Exploit Airflow Configuration & Deployment

  • Weak Credentials (DB Creds) [HIGH-RISK]:

    • Description: Airflow's connection to its metadata database uses weak or default credentials, allowing access to the database and potential modification of DAGs, tasks, or user accounts.
    • Likelihood: Low to Medium
    • Impact: Very High
    • Effort: Low
    • Skill Level: Novice
    • Detection Difficulty: Medium

  • Exposed API (Web UI API Key) [HIGH-RISK]:

    • Description: The Airflow web UI API is exposed to the internet without proper authentication or authorization, allowing attackers to control Airflow.
    • Likelihood: Low to Medium
    • Impact: Very High
    • Effort: Low
    • Skill Level: Novice
    • Detection Difficulty: Easy

  • Default Credentials (Admin PW) [CRITICAL]:

    • Description: The default Airflow administrator password is not changed, providing easy administrative access.
    • Likelihood: Low
    • Impact: Very High
    • Effort: Very Low
    • Skill Level: Novice
    • Detection Difficulty: Very Easy