This topic describes the properties of an OSS table.
auto.create.location
This property is used to automatically create an OSS directory and a table in this
directory in DLA. Example:
CREATE EXTERNAL TABLE person (
`id` int,
`name` string,
`age` int
)
STORED AS TEXTFILE
LOCATION 'oss://bucket001/dir001/'
TBLPROPERTIES (
'auto.create.location'='true'
);
compression.type
This property is used to manually specify the compression format. When DLA is used
to write data into OSS, the default compression format is snappy. Example:
CREATE EXTERNAL TABLE person (
`id` int,
`name` string,
`age` int
)
STORED AS TEXTFILE
LOCATION 'oss://bucket001/dir001/'
TBLPROPERTIES (
'compression.type'='gzip'
);
Note DLA supports the snappy and gzip compression formats. DLA also supports uncompressed data.
directory.odps
This property is used to use DLA to analyze OSS table data in MaxCompute. Example:
CREATE EXTERNAL TABLE person (
`id` int,
`name` string,
`age` int
)
STORED AS TEXTFILE
LOCATION 'oss://bucket001/dir001/'
TBLPROPERTIES (
'directory.odps'='true'
);
Note The directory structure of a MaxCompute external table is different from that of a
common OSS table. If you do not specify the
directory.odps
property, data files cannot be found.
skip.header.line.count
This property is used to specify the number of rows to be skipped when you analyze
text-type data in DLA. For example, if the first row of a CSV file is a file header,
this row must be skipped during data analysis. Example:
CREATE EXTERNAL TABLE person (
`id` int,
`name` string,
`age` int
)
STORED AS TEXTFILE
LOCATION 'oss://bucket001/dir001/'
TBLPROPERTIES ("skip.header.line.count"="1");
orc.use-column-names
This property is used to associate data of the ORC type with metadata based on column
names in an ORC file instead of the sequence of columns. Example:
CREATE EXTERNAL TABLE person (
`id` int,
`name` string,
`age` int
)
STORED AS TEXTFILE
LOCATION 'oss://bucket001/dir001/'
TBLPROPERTIES ("orc.use-column-names"="true");