Inside a Parquet file: how row groups, column chunks, and pages are laid out, what footer metadata stores, and how their sizes affect query performance.
How Parquet handles schema evolution when you add, remove, retype, rename, or reorder columns, how it compares to other formats, and best practices.
The compression codecs Parquet supports (Snappy, Gzip, Brotli, Zstandard, LZO), the speed and size tradeoffs of each, and how they pair with encoding.
How Parquet encodings work and when to use them: dictionary, run-length, bit-packing, delta, and plain, and how encoding pairs with compression.
