You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+5-2Lines changed: 5 additions & 2 deletions
Original file line number
Diff line number
Diff line change
@@ -23,10 +23,13 @@ A File Output Plugin for Embulk to write HDFS.
23
23
-**overwrite** overwrite files when the same filenames already exists (boolean, default: `false`)
24
24
-*caution*: even if this property is `true`, this does not mean ensuring the idempotence. if you want to ensure the idempotence, you need the procedures to remove output files after or before running.
25
25
-**doas** username which access to Hdfs (string, default: executed user)
26
-
-**delete_in_advance** delete files and directories having `path_prefix` in advance (boolean, default: `false`)
26
+
-**delete_in_advance** delete files and directories having `path_prefix` in advance (enum, default: `NONE`)
27
+
-`NONE`: do nothing
28
+
-`FILE_ONLY`: delete files
29
+
-`RECURSIVE`: delete files and directories
27
30
28
31
## CAUTION
29
-
If you use `hadoop` user (hdfs admin user) as `doas`, and if `delete_in_advance` is true,
32
+
If you use `hadoop` user (hdfs admin user) as `doas`, and if `delete_in_advance` is `RECURSIVE`,
30
33
`embulk-output-hdfs` can delete any files and directories you indicate as `path_prefix`,
31
34
this means `embulk-output-hdfs` can destroy your hdfs.
32
35
So, please be careful when you use `delete_in_advance` option and `doas` option ...
0 commit comments