How sonali2109/skycrate works
Project type and purpose
• SkyCrate is a full-stack web-based file management system built around a Spring Boot backend and a Hadoop (HDFS) storage layer.
• The system exposes REST endpoints for file operations, authentication, and HDFS interactions, while the frontend (not in this repo) consumes these APIs and provides i18n support.
File structure
• Top-level Dockerfile, Maven wrapper, pom.xml → containerized Spring Boot microservice.
• src/main/java/com/skycrate/backend/…
– config/ → Spring configuration beans (security, HDFS, caching, HTTPS redirect).
– controller/ → REST controllers (AuthController, FileController, HDFScontroller).
– dto/ → request & response payload models with validation.
– entity/ → JPA entities (User, FileMetadata, RefreshToken).
– repository/ → Spring Data JPA repos.
– security/ → JWT filter, token blacklist, encryption service.
– services/ → business logic (AuthenticationService, FileService, HDFSOperations, RateLimiterService, etc.).
– utils/ → key generation & encryption utilities.
• src/main/resources/ → application configs, prod profile, keystore.
README for additional insights
• Dockerized Hadoop cluster behind Tailscale → persistent HDFS service
• Hybrid RSA-AES file encryption per user
• JWT auth with refresh tokens + blacklist on logout
• Brute-force protection via in-code rate limiting
• Automatic HDFS directory provisioning on user signup
• Audit logging & DTO validation
Detail without over-fragmentation
• Major sub-systems (config, security, business, data).
• Unique features: hybrid encryption, automated HDFS provisioning, rate limiting, internationalization entry point.