This repository has been archived by the owner on Dec 8, 2021. It is now read-only.
Releases: pingcap/tidb-lightning
Releases · pingcap/tidb-lightning
v4.0.11
Improvements
- Create tables in parallel to improve import performance #502
- Skip splitting Regions to improve import performance if the engine's total size is smaller than the Region size #524
- Add a import progress bar and optimize the accuracy of restore progress #506
Bug Fixes
- Fix a bug that
column count mismatch
might be triggered when using the TiDB-backend #535 - Fix a bug that TiDB-backend panics if the column counts of the source file and the column counts of the target table mismatch #528
- Fix a bug that TiKV might unexpectedly panic during TiDB Lightning's data import #554
v4.0.10
Improvements
- Support retrying S3 reads on error #533
Bug Fixes
- Fix a bug that the retry scan region key is not encoded, which causes the retried Region is not split #531
- Fix a bug that lost create table error during multiple tables created #530
- Fix the bug that lightning will trim all the empty sep when trim-last-sep is true, which causes "column count mismatch" in tidb backend #535
v5.0.0-rc
v4.0.9
Improvements
- Filter out all system schemas by default #459
- Support setting a default value for the auto-random primary key for the Local-backend or Importer-backend #457
- Use range properties to make the range split more precise in Local-backend #422
- Support a human-readable format (such as "2.5 GiB") in
tikv-importer.region-split-size
,mydumper.read-block-size
,mydumper.batch-size
, andmydumper.max-region-size
#471