Skip to content

Latest commit

 

History

History
40 lines (30 loc) · 2.02 KB

release-pre-ga.md

File metadata and controls

40 lines (30 loc) · 2.02 KB
title aliases summary
Pre-GA release notes
/docs/dev/releases/release-pre-ga/
/docs/dev/releases/prega/
TiDB Pre-GA release on August 30, 2017, focuses on MySQL compatibility, SQL optimization, stability, and performance. TiDB introduces SQL query optimizer enhancements, MySQL compatibility, JSON type support, and memory consumption reduction. Placement Driver (PD) now supports manual leader change, while TiKV uses dedicated Rocksdb for Raft log storage and improves performance. TiDB Connector for Spark Beta Release implements predicates pushdown, aggregation pushdown, and range pruning, capable of running TPC+H queries.

Pre-GA Release Notes

On August 30, 2017, TiDB Pre-GA is released! This release is focused on MySQL compatibility, SQL optimization, stability, and performance.

TiDB

  • The SQL query optimizer:
    • Adjust the cost model
    • Use index scan to handle the where clause with the compare expression which has different types on each side
    • Support the Greedy algorithm based Join Reorder
  • Many enhancements have been introduced to be more compatible with MySQL
  • Support Natural Join
  • Support the JSON type (Experimental), including the query, update and index of the JSON fields
  • Prune the useless data to reduce the consumption of the executor memory
  • Support configuring prioritization in the SQL statements and automatically set the prioritization for some of the statements according to the query type
  • Completed the expression refactor and the speed is increased by about 30%

Placement Driver (PD)

  • Support manually changing the leader of the PD cluster

TiKV

  • Use dedicated Rocksdb instance to store Raft log
  • Use DeleteRange to speed up the deleting of replicas
  • Coprocessor now supports more pushdown operators
  • Improve the performance and stability

TiDB Connector for Spark Beta Release

  • Implement the predicates pushdown
  • Implement the aggregation pushdown
  • Implement range pruning
  • Capable of running full set of TPC+H except for one query that needs view support