BinaryDoc generates Documents with rich diagrams from Binary files directly for Java.

This docker image needs to work together with mysql database with specified mysql.cnf configurations. An sample docker-compose.yml file has been provided.

Sample Diagrams in Generated documents:

• Control Flow Graph for Java method bytecode

• UML Sequence Diagram for Java method bytecode

• UML Hierarchy Diagram for Java Class

• Dependency Network diagram for Java package

• Jar,Jmod files Dependency Network Diagram (sfdp layout)

There are web sites containing the documentes and diagrams generated by BinaryDoc:

• BinaryDoc for OpenJDK
• BinaryDoc for Android
• BinaryDoc for Kotlin
• BinaryDoc for Groovy
• BinaryDoc for Scala
• BinaryDoc for Spring framework
• BinaryDoc for Apache Cassandra
• BinaryDoc for Apache Tomcat
• BinaryDoc for Eclipse Jetty

This is an sample docker compose setup for BinaryDoc, which generates Documents with rich diagrams from Binary files directly for Java.

• docker-compose.yml

Warning

• This docker compose setup is designed for none-production environment, better for evaluation or testing box
• Password is set in .env, while password vault could be used in production
• Default RAM in mysql.cnf is small, while 128 GB or more RAM is available in production

BinaryDoc is a compute-intensive software which needs powerful hardware.

Here we list the suggested minimal hardware requirements:

• CPU: 8 or more threads
• RAM: 24 GB minimal, 64 GB is suggested for better performance, based on the size of the Java application to be parsed
• Disk type: SSD Drive is suggested, M.2 SSD or NVMe SSD are preferred
• Disk size: 20 GB or more avialbe disk space is suggested
  • The size is based on the binary data to be parsed.
  • Example: OpenJDK 11 binary data size is 387.9 MB, and the current corresponding DB size is 7 GB.

Customize the config files

• .env
  • Change the DB password (default=123456) when needed
  • Change the TCP ports when needed
• mysql.cnf
  • Change innodb_buffer_pool_size based on avabile RAM hardware and the size of application to be parsed
  • Change innodb_buffer_pool_instances based on parallel workloads

If running Docker on Windows

• If running on other system, ignore this step
• Change the file etc/mysql/conf.d/mysql.cnf permission as read only, or else the mysql file will not load it

Start the Docker Instances

• docker-compose up -d

Run the Parser for current OpenJDK for demo:

• ./run-parse-demo.sh
  • Note. the parser execution takes about 1 hour on M.2 SSD disk, and needs up to 24 GB RAM

Run the Parser for your application:

• Put your Java Application to be Parsed to the app-to-parse folder
  • Example: Copy the application .jmod or .jar files to this folder
• Start the parser
  • ./run-parse-app.sh

Access the Web Site

• http://127.0.0.1:10180/

(Optional) Access the Backend DB if interested

• http://127.0.0.1:10190/
  • System: MySQL
  • Server: binarydoc-db
  • Username: root
  • Password: 123456, or the password set in the .env file
  • Database: binarydocjvmadm
• MySQL Workbench, or Other DB Access Tools
  • Server: localhost
  • Port: 13306

When we want to Delete Existing Docker Instances, and to Start over again

• ./rebuild.sh

When we want login to the Running Docker container

• sudo docker-compose exec binarydoc /bin/bash

TCP Ports Mapping

• We may change the ports 10180, 10190, 13306 in .env when necessary

• Support: Report Issues in Github
• License: Apache License 2.0