ZhouJinsong

a subtask of #38

Hive parquet reader/writer have different impelemention from iceberg，like: - use INT96 to store timestamp value So we need finished this to support #38

The transaction TAG of table on AMS is expected to list transaction which user commit. But now it list transaction optimizer produce either.

Both base store and chage store are iceberg table for now. This issue propose a hive table base store implemention， proposal link: https://as9e2kjei0.feishu.cn/docs/doccn5L4ZLTfvIlcElDbeVFvt8d

To support original write and alter to hive table， we need a meta synchronizer to sync meta change and data change to iceberg table in base store, subtask of #38

A subtask of #38

A subtask of #38

### Search before asking - [X] I have searched in the [issues](https://github.com/NetEase/arctic/issues?page=4&q=is%3Aissue) and found no similar issues. ### What would you like to be improved? Arctic create catalog with authrization...

[Feature]: Support insert part of columns of mixed-format table with flink in streaming mode

3

### Description Support insert part of columns of arctic table with flink in streaming mode ### Use case/motivation Arctic only support insert all table columns with flink in streaming mode...

### Search before asking - [X] I have searched in the [issues](https://github.com/NetEase/arctic/issues?page=4&q=is%3Aissue) and found no similar issues. ### What would you like to be improved? Arctic support hadoop and hive...