jasonz-lance
f5e23ab080
Two independent fixes to 2-phase shuffled reads, found by the same gatekeeper review pass: - Packed checkpoint resume: the 2-phase branch of _iter_owned read the (empty, in packed mode) row-mode cursor maps instead of _pack_consumed, so restoring a packed checkpoint silently replayed already-consumed documents. Now mirrors the 1-phase branch's check. - Version pinning + remote (Cloud) table support: 2-phase previously called self._table.to_lance() fresh at iteration time, which (a) raises NotImplementedError on RemoteTable, since to_lance() only works against a local table, and (b) re-reads whatever version is live at that moment rather than the one the block plan was computed against, so a write to the table between construction and iteration (or between epochs) could shift which rows a block refers to. Now pins table.version (and rejects a table checked out to a non-default branch, which isn't supported yet) at construction, and reads through a private handle reopened via the same connection_factory/pickle-state machinery already used for worker reconnects, checked out to that pinned version. Reads go through Table's backend-agnostic query builder (search/select/take_offsets), the same interface RemoteTable implements, instead of pylance's to_lance(). take_offsets doesn't guarantee its output order, so filtered block reads always additionally select "_rowoffset" and sort locally by it to restore the requested order. Extends MockPermutationServer (utils.py) to answer the offset/limit scan and _rowoffset-filtered take request shapes 2-phase now sends, and adds tests for both fixes plus 2-phase over a (mocked) remote table, unfiltered and filtered+shuffled.
LanceDB Python SDK
A Python library for LanceDB.
Installation
pip install lancedb
Pre-Haswell x86_64 hosts: lancedb-compat
The default lancedb wheel targets x86-64-haswell (AVX2 + FMA + F16C) for full performance on modern hardware. Pre-Haswell hosts — Intel Sandy Bridge / Ivy Bridge / Westmere; AMD Bulldozer / Piledriver / Steamroller — don't have AVX2 and crash with Illegal instruction at import lancedb.
For those hosts, install the lancedb-compat package instead:
pip install lancedb-compat
Same Python API (import lancedb works as usual). The compat wheel is compiled at the x86-64-v2 baseline (Nehalem-class) and uses runtime SIMD dispatch in the embedded lance crate to pick the right kernel tier (scalar / AVX / AVX+FMA / AVX2+FMA / AVX-512) at load time, so it still goes fast on modern hardware while running cleanly on the pre-Haswell silicon. Use lance.simd_info() from Python to verify which tier was selected.
lancedb and lancedb-compat install to the same lancedb/ namespace and conflict at install time. Pick one. To switch, pip uninstall lancedb first, then pip install lancedb-compat (or vice-versa).
If you need a custom baseline (or lancedb-compat isn't yet published for your platform), build from source with the override:
RUSTFLAGS="-C target-cpu=x86-64-v2" maturin build --release
pip install ./target/wheels/lancedb-*.whl
Preview Releases
Stable releases are created about every 2 weeks. For the latest features and bug fixes, you can install the preview release. These releases receive the same level of testing as stable releases, but are not guaranteed to be available for more than 6 months after they are released. Once your application is stable, we recommend switching to stable releases.
pip install --pre --extra-index-url https://pypi.fury.io/lancedb/ lancedb
Threading in CPU-limited containers
LanceDB uses separate pools for compute work and storage I/O. On a container with two visible CPUs, current releases intentionally use one compute worker by default; no manual configuration is needed. If every query logs an I/O core reservation warning on a two-CPU container, upgrade from LanceDB 0.21.1 or earlier.
The two commonly tuned environment variables control different resources:
LANCE_CPU_THREADSoverrides the number of compute workers. One worker is the appropriate setting for a two-CPU container when an explicit override is needed.LANCE_IO_THREADScontrols concurrent storage operations, not reserved CPU cores. Its default can be greater than the number of CPUs because I/O workers spend much of their time waiting for storage.
Keep the defaults unless measurements show that the workload benefits from an override. See the Lance threading model for the current defaults and tuning guidance.
Usage
Basic Example
import lancedb
db = lancedb.connect('<PATH_TO_LANCEDB_DATASET>')
table = db.open_table('my_table')
results = table.search([0.1, 0.3]).limit(20).to_list()
print(results)
Development
See CONTRIBUTING.md for information on how to contribute to LanceDB.