mirror of
https://github.com/qdrant/qdrant.git
synced 2026-08-06 10:00:58 -05:00
fa3d756914
UniversalReadFileOps mixed read-side operations (from_context, list_files, exists) with mutating ones (create, create_dir, remove, remove_dir, atomic_save). Move the mutating operations to a new UniversalWriteFileOps subtrait, mirroring UniversalRead/UniversalWrite. - UniversalWrite now requires Fs: UniversalWriteFileOps, so generic consumers (gridstore) keep reaching write ops through S::Fs. - ReadOnlyFs drops its runtime-erroring write stubs: read-only is now a compile-time property. - DiskCacheFs implements the write side only when the remote fs does. - BlobFs and the async AsyncRead trait are read-only: the async write methods are removed from AsyncRead and its backends (object store, uio-grpc) along with the tests that exercised them. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
gridstore
New storage for variable-sized values using mmap.
Design
- IDs are sequential integers, starting at 0.
- The storage is divided into file pages of fixed size (32MB).
- Data can be written and read across multiple pages.
- Those pages are mapped into memory using mmap.
- Data units are blocks of fixed size (128 bytes).
- Values span an integer number of contiguous blocks.
- Values are compressed with lz4
- Each block is mapped to a bit in the bitmask.
- A region is a fixed number of contiguous blocks.
- Gaps of free blocks in each region are tracked in a file.
- Deletes mark the block as deleted (in-memory) & updates their region
- Updates:
- not done in place, always a new value is inserted
- calculation of the new regions gaps is done on the fly
- the tracker is updated in-memory, only persisted on flush
- Supports multiple threads reading and single thread writing
- One file per page, one file for tracker, one file for bitmask, and one file for gaps.
TODOs
- dictionary compression to optimize payload key repetition
- validate the usage with a block storage via HTTP range requests