[FLINK-18802] Create an uber jar when packaging flink-avro for sql client

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.12.0
• Component/s: Formats (JSON, Avro, Parquet, ORC, SequenceFile)
• Labels: pull-request-available
• Release Note: The sql client jar was renamed to flink-sql-avro-${flink.version}.jar, previously flink-avro-${flink.version}-sql-jar.jar. Moreover it is no longer needed to add avro dependencies manually.
• Assignee: Dawid Wysakowicz
• Reporter: Qingsheng Ren
• Created: 03/Aug/20 09:33
• Updated: 16/Sep/20 06:40
• Resolved: 14/Sep/20 19:49

Description

Currently users have to provide dependencies such as avro, jackson-core-asl, jackson-mapper-asl and joda-time in the job jar for DataStream jobs, or manually copy them into flink/lib in SQL jobs when using avro formatting.

It's better to generate an uber jar including these dependencies when packaging flink-avro module. 

Issue Links

• is related to: FLINK-18192 Upgrade to Avro version 1.10.0 from 1.8.2
• supercedes: FLINK-19100 Fix note about hadoop dependency in flink-avro
• links to: GitHub Pull Request #13072
• links to: GitHub Pull Request #13373

Sub-Tasks

• Update the Chinese documentation: Closed
• Update the flink-web: Closed