陈磊 created HUDI-7941:
------------------------
Summary: add show_file_status procedure
Key: HUDI-7941
URL: https://issues.apache.org/jira/browse/HUDI-7941
Project: Apache Hudi
Issue Type: New Feature
Reporter: 陈磊
When incrementally consuming the hudi table or performing clustering or
compaction operations on the hudi table, it is often found that a certain file
does not exist. If you want to know which operation deleted the file, it is a
very troublesome operation. For this purpose, we provide a tool
`show_file_status` to view whether a specified file has been deleted and what
actions have been taken to delete it.
usage:
call show_file_status(table => '$tableName', partition => '$partition', file =>
'$fileName')
call show_file_status(table => '$tableName', file => '$fileName')
output:
1）the file was deleted by the restore action
+-------+-------+-----------------+--------+---------+
|status |action |instant |timeline|full_path|
+-------+-------+-----------------+--------+---------+
|deleted|restore|20240629225539880|active | |
+-------+-------+-----------------+--------+---------+
2）the file has been deleted in other ways, such as hdfs dfs -rm
+-------+------+-------+--------+---------+
|status |action|instant|timeline|full_path|
+-------+------+-------+--------+---------+
|unknown| | | | |
+-------+------+-------+--------+---------+
3) the file exists
+------+------+-------+--------+---------------------------------------------------------------------------------------------------------------------------------------------------------------+
|status|action|instant|timeline|full_path
|
+------+------+-------+--------+---------------------------------------------------------------------------------------------------------------------------------------------------------------+
|exist | | |active
|/Users/xx/xx/others/data/hudi-warehouse/source1/hudi_mor_append/sex=0/85ad0f44-22bf-4733-99bf-06382d6eacd5-0_0-130-89_20240629230123162.parquet|
+------+------+-------+--------+---------------------------------------------------------------------------------------------------------------------------------------------------------------+
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
