doris spark同步到doris

Last updated on September 15, 2024 pm

🧙 Questions

将spark中的数据导入doris中

☄️ Ideas

docker exec -it isxcode-mysql bash
mysql -u root -pispong123 -h isxcode -P 30131
SHOW PROC ‘/brokers’;

Note:
RESOURCE NAME 不能使用 -

CREATE EXTERNAL RESOURCE "ispong_spark"
PROPERTIES
(
  "type" = "spark",
  "spark.master" = "yarn",
  "spark.submit.deployMode" = "cluster",
  -- "spark.jars" = "/data/cdh/cloudera/parcels/CDH-6.2.0-1.cdh6.2.0.p0.967373/lib/*",
  -- "spark.files" = "/tmp/aaa,/tmp/bbb",
  "spark.executor.memory" = "1g",
  -- "spark.yarn.queue" = "ispong-queue",
  "spark.hadoop.yarn.resourcemanager.address" = "172.23.39.206:8032",
  "spark.hadoop.fs.defaultFS" = "hdfs://172.23.39.206:30116",
  "working_dir" = "hdfs://172.23.39.206:30116/tmp/doris",
  "broker" = "broker_ispong",
  "broker.username" = "",
  "broker.password" = ""
);
SHOW RESOURCES;
SHOW PROC "/resources";
DROP RESOURCE ispong_spark;

CREATE EXTERNAL RESOURCE "ispong_spark"
PROPERTIES
(
  "type" = "spark",
  "spark.master" = "spark://172.23.39.206:7777",
  "spark.submit.deployMode" = "client",
  "working_dir" = "hdfs://172.23.39.206:30116/tmp/doris",
  "broker" = "broker_ispong"
);

CREATE EXTERNAL TABLE ispong_table_hive
(
    username varchar(50),
    age INT
)
ENGINE=hive
properties
(
  "database" = "cdh_dev", 
  "table" = "ispong_table",
  "hive.metastore.uris" = "thrift://172.23.39.206:30123"
);

需要将spark的jar引入到项目

vim /data/doris/fe/conf/fe.conf 

# === vim /data/doris/fe/conf/fe.conf  ===

spark_home_default_dir=/data/cdh/cloudera/parcels/CDH/lib/spark
spark_resource_patj=/data/doris/fe/lib/spark2x/jars/spark-2x.zip

# === vim /data/doris/fe/conf/fe.conf  ===

cd ~/spark-jars
zip spark-2x.zip ./*
mkdir -p /data/doris/fe/lib/spark2x/jars
mv spark-2x.zip /data/doris/fe/lib/spark2x/jars/
hadoop fs -mkdir -p /tmp/doris
hadoop fs -ls hdfs://172.23.39.206:30116/tmp/doris

LOAD LABEL ispong_db.ispong_job_10
(
    DATA INFILE("hdfs://isxcode:30116/user/hive/warehouse/cdh_dev.db/ispong_table/*")
    INTO TABLE ispong_table
    COLUMNS TERMINATED BY ","
    (username,age)
    SET
    (
        username=username,
        age=age
    )
)
WITH RESOURCE 'ispong_spark'
(
    "spark.executor.memory" = "2g",
    "spark.shuffle.compress" = "true"
)
PROPERTIES
(
    "timeout" = "3600"
);

ERROR 1064 (HY000): errCode = 2, detailMessage = Spark Load is coming soon

SHOW TABLES;
-- 导入数据
-- LOAD LABEL ispong_db.ispong_job_1
-- (
--     DATA INFILE("hdfs://isxcode:30116/user/hive/warehouse/cdh_dev.db/ispong_table/*")
--     INTO TABLE ispong_table_copy
--     COLUMNS TERMINATED BY ","
--     (username,age)
--     SET
--     (
--         username=username,
--         age=age
--     )
-- )
-- WITH BROKER 'broker_ispong'
-- (
--     "username"="",
--     "password"=""
-- )
-- PROPERTIES
-- (
--     "timeout" = "3600"
-- );
-- 查询load
SHOW LOAD;
-- 取消load
CANCEL LOAD FROM ispong_db WHERE LABEL = "ispong_label_2";
-- 删除label
DELETE FROM ispong_db WHERE LABEL = "ispong_label_2";
-- 查询表 
SELECT * FROM ispong_db.ispong_table;