[HIVE-13170] HiveAccumuloTableOutputFormat should implement HiveOutputFormat to ensure compatibility

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.2.1, 2.0.0
• Component/s: Accumulo Storage Handler
• Labels: pull-request-available
• Assignee: Teng Qiu
• Reporter: Teng Qiu
• Created: 26/Feb/16 13:57
• Updated: 24/Jun/20 00:28

Description

this issue was caused by same reason described in https://issues.apache.org/jira/browse/HIVE-11166

both HiveAccumuloTableOutputFormat and HiveHBaseTableOutputFormat does not implemented HiveOutputFormat, it may break the compatibility in some other APIs that are using hive, such as spark's API.

spark expects the OutputFormat called by hive storage handler is some kind of HiveOutputFormat. which is totally reasonable.

and since they are OutputFormat for hive storage handler, they should not only extend the 3rd party OutputFormat (AccumuloOutputFormat or hbase.TableOutputFormat), but also implement HiveOutputFormat interface.

Issue Links

• relates to: HIVE-11166 HiveHBaseTableOutputFormat can't call getFileExtension(JobConf jc, boolean isCompressed, HiveOutputFormat<?, ?> hiveOutputFormat)
• links to: GitHub Pull Request #66