Abstract
This paper reports a large scale and secure cloud file integrity which the implementation put forward to protect out hosted data against illegal changes, attacks, or system failures. The proposed system addresses the issue of present cloud storage which does not have real time verification. System use cryptographic hashing (SHA-256/SHA-512) with a chunk-based approach to manage large files which also sees to that do not have delays. At the time of upload, the system has reference hash and metadata in a secure way. The proposed system always does a periodic or on demand check by which it generates a new hash and compares it with the stored reference. The proposed system has real time alert features, secure audit trails, versioning, and support for many clouds (AWS S3, Azure Blob, Google Cloud Storage). The system leverages Python for hashing, Node.js for interfaces, and MongoDB for managing secure metadata. Experimental evaluation demonstrates $\mathbf{1 0 0 \%}$ detection accuracy without false positives. It managed to identify all the corrupted files and, as a result, it can send the notifications or the alerts immediately. The use of chunk-level hashing made the process of verification more efficient and it helped in exactly pointing out the regions which were tampered with.