Support the AES-GCM-CTR-V1 encryption algorithm for Parquet modular encryption
- Dominant language
- Rust
- Stars
- 3.6k
- Forks
- 1.3k
- Avg merge
- 2d 18h
- Merged PRs (30d)
- 169
Description
Follow up task to https://github.com/apache/arrow-rs/pull/6637, which adds read support for files that use Parquet modular encryption.
Currently only the AES-GCM-V1 algorithm is supported. AES-GCM-CTR-V1 requires using un-authenticated AES-CTR for data pages, which isn't yet supported by the Ring library.
There's a Ring issue open for this and the maintainer has expressed that they'd be open to accepting this feature but it isn't planned: https://github.com/briansmith/ring/issues/414
For details on the two algorithms, see https://github.com/apache/parquet-format/blob/master/Encryption.md#42-parquet-encryption-algorithms.
Contributor guide
Research direction
Start by reviewing PR 6637's Parquet modular-encryption read support and the Parquet Encryption.md section linked in the issue. Check Ring issue 414 to understand the upstream AES-CTR limitation and whether the required support is available. Done means arrow-rs can read files using AES-GCM-CTR-V1 as well as AES-GCM-V1.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- cryptography
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100