phoenix 利用CsvBulkLoadTool 批量带入数据并自动创建索引

Posted 2020-06-14

需要先创建表：

CREATE TABLE IF NOT EXISTS population (
  state CHAR(2) NOT NULL,  city VARCHAR NOT NULL,  population BIGINT 
 CONSTRAINT my_pk PRIMARY KEY (state, city));

在phoenix 目录下执行

hadoop jar /home/phoenix-4.6.0-HBase-1.0-bin/phoenix-4.6.0-HBase-1.0-client.jar   org.apache.phoenix.mapreduce.CsvBulkLoadTool -t POPULATION -i /datas/us_population.csv

-t ：tableName

-i: input file 文件必须在hdfs文件上。

后查询表数据是空。

问题： ERROR mapreduce.CsvBulkLoadTool: Error Wrong FS: file:/home/hadoop/tmp/partitions_101bd67a-ec2c-4808-bc9f-bf4cd6ea74b9, expected: hdfs://node11:9000 occurred submitting CSVBulkLoad 

 

 参数意义：

-a,--array-delimiter <arg> Array element delimiter (optional)
-c,--import-columns <arg> Comma-separated list of columns to be
imported
-d,--delimiter <arg> Input delimiter, defaults to comma
-e,--escape <arg> Supply a custom escape character, default is
a backslash
-g,--ignore-errors Ignore input errors
-h,--help Show this help and quit
-i,--input <arg> Input CSV path (mandatory)
-it,--index-table <arg> Phoenix index table name when just loading
this particualar index table
-o,--output <arg> Output path for temporary HFiles (optional)
-q,--quote <arg> Supply a custom phrase delimiter, defaults
to double quote character
-s,--schema <arg> Phoenix schema name (optional)
-t,--table <arg> Phoenix table name (mandatory)
-z,--zookeeper <arg> Supply zookeeper connection details
(optional)