Dingodb Dingo Versions Save

A multi-modal vector database that supports upserts and vector queries using unified SQL (MySQL-Compatible) on structured and unstructured data, while meeting the requirements of high concurrency and ultra-low latency.

dingo-v0.1.0

1 year ago

DingoDB 0.1.0 Release Notes

  • Cluster
    1. Distributed computing. Cluster nodes are classified into coordinator role and executor role.
    2. Distributed meta data storage. Support creating and dropping meta data of tables.
    3. Coordinators support SQL parsing and optimizing, job creating and distributing, result collecting.
    4. Executors support task executing.
  • Data store
    1. Using RocksDB storage.
    2. Encoding and decoding in Apache Avro format.
    3. Table partitioning by hash of primary columns.
  • SQL parsing and executing
    1. Create (CREATE TABLE) and drop table (DROP TABLE).
    2. Supporting common SQL data types: TINYINT, INT, BIGINT, CHAR, VARCHAR, FLOAT, DOUBLE, BOOLEAN
    3. Insert into (INSERT INTO TABLE) and delete from (DELETE FROM TABLE) table.
    4. Query table (SELECT).
    5. Support filtering and projecting in query.
    6. Support expressions in filter conditions and projecting columns.
    7. Support point query.
  • User interface
    1. Command line interface (CLI)
    2. Support SQL input and executing in CLI.
    3. Output query results in table format in CLI.
    4. Output time consumed by query in CLI.

v0.1.0

1 year ago

DingoDB 0.1.0 Release Notes

  • Cluster
    1. Distributed computing. Cluster nodes are classified into coordinator role and executor role.
    2. Distributed meta data storage. Support creating and dropping meta data of tables.
    3. Coordinators support SQL parsing and optimizing, job creating and distributing, result collecting.
    4. Executors support task executing.
  • Data store
    1. Using RocksDB storage.
    2. Encoding and decoding in Apache Avro format.
    3. Table partitioning by hash of primary columns.
  • SQL parsing and executing
    1. Create (CREATE TABLE) and drop table (DROP TABLE).
    2. Supporting common SQL data types: TINYINT, INT, BIGINT, CHAR, VARCHAR, FLOAT, DOUBLE, BOOLEAN
    3. Insert into (INSERT INTO TABLE) and delete from (DELETE FROM TABLE) table.
    4. Query table (SELECT).
    5. Support filtering and projecting in query.
    6. Support expressions in filter conditions and projecting columns.
    7. Support point query.
  • User interface
    1. Command line interface (CLI)
    2. Support SQL input and executing in CLI.
    3. Output query results in table format in CLI.
    4. Output time consumed by query in CLI.

dingo-v0.2.0

1 year ago

  • Architecture

    1. Refactor DingoDB architecture abandon Zookeeper, Kafka and Helix.
    2. Using raft as the consensus protocol to make agreement across multiple nodes on membership selection and data replication.
    3. Region is proposed as the unit of data replication, it can be scheduled, split, managed by coordinator.
    4. The distributed file system is replaced by distributed key-value implemented by raft and rocksdb.

  • Distributed Storage

    1. Support region to replicate across multiple nodes.
    2. Support region to split based on policies such as key counts or region size.
    3. Support Region to perform periodic snapshot.

  • SQL

    1. Support more aggregation functions, such as min,max,avg, etc.
    2. Support insert into ... select.

  • Client Tools

    1. thin jdbc driver

v0.2.0

1 year ago

  • Architecture

    1. Refactor DingoDB architecture abandon Zookeeper, Kafka and Helix.
    2. Using raft as the consensus protocol to make agreement across multiple nodes on membership selection and data replication.
    3. Region is proposed as the unit of data replication, it can be scheduled, split, managed by coordinator.
    4. The distributed file system is replaced by distributed key-value implemented by raft and rocksdb.

  • Distributed Storage

    1. Support region to replicate across multiple nodes.
    2. Support region to split based on policies such as key counts or region size.
    3. Support Region to perform periodic snapshot.

  • SQL

    1. Support more aggregation functions, such as min,max,avg, etc.
    2. Support insert into ... select.

  • Client Tools

    1. thin jdbc driver

dingov0.1.0

2 years ago

DingoDB 0.1.0 Release Notes

  • Cluster
    1. Distributed computing. Cluster nodes are classified into coordinator role and executor role.
    2. Distributed meta data storage. Support creating and dropping meta data of tables.
    3. Coordinators support SQL parsing and optimizing, job creating and distributing, result collecting.
    4. Executors support task executing.
  • Data store
    1. Using RocksDB storage.
    2. Encoding and decoding in Apache Avro format.
    3. Table partitioning by hash of primary columns.
  • SQL parsing and executing
    1. Create (CREATE TABLE) and drop table (DROP TABLE).
    2. Supporting common SQL data types: TINYINT, INT, BIGINT, CHAR, VARCHAR, FLOAT, DOUBLE, BOOLEAN
    3. Insert into (INSERT INTO TABLE) and delete from (DELETE FROM TABLE) table.
    4. Query table (SELECT).
    5. Support filtering and projecting in query.
    6. Support expressions in filter conditions and projecting columns.
    7. Support point query.
  • User interface
    1. Command line interface (CLI)
    2. Support SQL input and executing in CLI.
    3. Output query results in table format in CLI.
    4. Output time consumed by query in CLI.