HDF5
A hierarchical data format commonly used for scientific computing and machine learning datasets. It supports an external storage mechanism that allows datasets to reference other files on disk, and carries risks of path traversal and file read vulnerabilities.
Timeline (last 90 days)
The HDF5 external storage mechanism allows datasets to reference other files on disk via relative or absolute paths, and when no path validation is performed it can be used for path traversal/file read attacks.
The project uses HDF5's native export format to store 500,000 rows of multi-dimensional time-series data, providing engineering-level guarantees in storage efficiency and access performance compared to CSV or JSON.
HDF5 is a file format developed by the National Center for Supercomputing Applications (NCSA) and maintained by the HDF Group, supporting hierarchical groups and dataset structures, multi-dimensional array storage, gzip/LZF compression, and chunked storage.
All Facts (2)
The HDF5 external storage mechanism allows datasets to reference other files on disk via relative or absolute paths, and when no path validation is performed it can be used for path traversal/file read attacks.
50%UnverifiedHDF5 is a file format developed by the National Center for Supercomputing Applications (NCSA) and maintained by the HDF Group, supporting hierarchical groups and dataset structures, multi-dimensional array storage, gzip/LZF compression, and chunked storage.
50%