Lakelet

Lakelet is a single-binary SQL engine for LakeHouse tables. No Spark cluster is needed, query data lake tables directly with SQL.

show catalogs;
use catalog hms;
select * from default.orders limit 10;

Features

Area	Support
Catalog	Hive Metastore, AWS Glue, Paimon filesystem, Iceberg REST
Table format	Iceberg, Hive, Paimon, Delta Lake
File format	Parquet, Hive TextFile
Storage	S3 / S3-compatible storage, Aliyun OSS, HDFS

Quick Start

Install the latest nightly binary into the current directory (Linux and macOS; Windows builds are on the releases page):

curl -fsSL https://lakelet.dev/install.sh | sh

Or build from source, more details see the docs:

cargo build --release
cp target/release/lakelet .
cp config_demo.toml config.toml

Configure at least one external catalog in config.toml:

[[catalog.hms]]
name = "hms"
metastore-uri = "127.0.0.1:9083"
s3-storage = { region = "us-east-1", endpoint = "http://127.0.0.1:9000", access-key = "admin", secret-key = "password", path-style-access = true }

Start the SQL shell:

./lakelet --config config.toml

Run one SQL command and exit:

./lakelet --config config.toml --command "show catalogs;"

Run SQL from a file:

./lakelet --config config.toml --file query.sql

Start the server:

./lakelet --config config.toml --server

Basic SQL

show catalogs;
use catalog hms;
show schemas;
use default;
show tables;
select * from my_table limit 10;

You can also use fully qualified table names:

select * from hms.default.my_table limit 10;

Documentation

https://lakelet.dev/