FROM: Makes the statement more readable.

URL: A URL accessible with a HTTP GET request.

URI: Available as of MySQL 9.4.0. The file location of the URI.

S3: The AWS S3 file location.

This requires the user privilege LOAD_FROM_S3.

See: Section 11.2.1.2, “Bulk Ingest Data to MySQL Server Limitations”.

COUNT: The number of files in a series of files.

For COUNT 5 and file_prefix set to data.csv., the five files would be: data.csv.1, data.csv.2, data.csv.3, data.csv.4, and data.csv.5.

IN PRIMARY KEY ORDER: Use when the data is already sorted. The values should be in ascending order within the file.

For a file series, the primary keys in each file must be disjoint and in ascending order from one file to the next.

PARALLEL: The number of concurrent threads to use. A typical value might be 16, 32 or 48. The default value is 16.

PARALLEL does not require CONCURRENT.

MEMORY: The amount of memory to use. A typical value might be 512M or 4G. The default value is 1G.

ALGORITHM: Set to BULK for bulk load. The file format is CSV.

COMPRESSION: The file compression algorithm. Bulk load supports the ZSTD algorithm.

options is a JSON object literal that includes:

An example that loads unsorted data from AWS S3 with 48 concurrent threads and 4G of memory:

mysql> GRANT LOAD_FROM_S3 ON *.* TO load_user@localhost;

mysql> LOAD DATA FROM S3 's3-us-east-1://innodb-bulkload-dev-1/lineitem.tbl'
        INTO TABLE lineitem
        FIELDS TERMINATED BY "|"
        OPTIONALLY ENCLOSED BY '"'
        LINES TERMINATED BY '\n'
        PARALLEL = 48
        MEMORY = 4G
        ALGORITHM=BULK;

An example that loads eight files of sorted data from AWS S3. The file_prefix ends with a period. The files are lineitem.tbl.1, lineitem.tbl.2, ... lineitem.tbl.8:

mysql> GRANT LOAD_FROM_S3 ON *.* TO load_user@localhost;

mysql> LOAD DATA FROM S3 's3-us-east-1://innodb-bulkload-dev-1/lineitem.tbl.' COUNT 8
        IN PRIMARY KEY ORDER
        INTO TABLE lineitem
        FIELDS TERMINATED BY "|"
        OPTIONALLY ENCLOSED BY '"'
        LINES TERMINATED BY '\n'
        ALGORITHM=BULK;

An example that performs a dry run on a sequence of MySQL Shell dump files compressed with the ZSTD algorithm:

mysql> GRANT LOAD_FROM_URL ON *.* TO load_user@localhost;

mysql> LOAD DATA FROM URL 
        '{"url-prefix","https://example.com/bucket/test@lineitem@","url-sequence-start",0,"url-suffix",".tsv.zst","url-prefix-last-append","@","is-dryrun",true}'
        COUNT 20
        INTO TABLE lineitem
        CHARACTER SET ???? COMPRESSION = {'ZSTD'}
        FIELDS TERMINATED BY "|"
        OPTIONALLY ENCLOSED BY '"'
        LINES TERMINATED BY '\n'
        IGNORE 20000 LINES
        ALGORITHM=BULK;

An example that loads data with the URI keyword (supported as of MySQL 9.4.0):

mysql> GRANT LOAD_FROM_URL ON *.* TO load_user@localhost;
  
mysql> LOAD DATA FROM URI 'https://data_files.com/data_files_1.tbl'
        INTO TABLE lineitem
        FIELDS TERMINATED BY "|"
        OPTIONALLY ENCLOSED BY '"'
        LINES TERMINATED BY '\n'
        ALGORITHM=BULK;

An example that monitors bulk load progress in a second session.