如何使用ojdbc14.jar在spark-sql-2.4.1v中将日期/时间戳作为lowerBound/upperBound传递？

问题描述 投票：0回答：3

我正在使用spark-sql-2.4.1v和ojdbc6.jar从oracle读取数据。

有如下oracle表

create table schema1.modal_vals(
  FAMILY_ID          NOT NULL NUMBER,
  INSERTION_DATE     NOT NULL DATE,
  ITEM_VALUE         VARCHAR2(4000),
  YEAR               NUMBER,
  QUARTER            NUMBER,
  LAST_UPDATE_DATE   DATE
)

加载样本数据：

insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-02","bbb-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-13","b+",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-17","bbb-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-13","bb",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-02","ccc-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-13","aa-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-OCT-13","a-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-03","bbb-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-13","b",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-FEB-03","aa+",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-13","aa+",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JAN-19","aaa+",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"30-JUN-18","ccc-",2013,2,null);
insert into modal_vals(FAMILY_ID,INSERTION_DATE,ITEM_VALUE,YEAR,QUARTER,LAST_UPDATE_DATE) values(2,"01-MAY-19","bb-",2013,2,null);

尝试将数据加载到spark-sql中，如下所示：

//please fill the respected oracle details

 DataFrameReader ora_df_reader  = spark.read().format("jdbc")
            .option("url", o_url)  
            .option("driver", Constants.ORACLE_DRIVER)
            .option("user", o_userName)
            .option("password", o_passwd)
            .option("fetchsize",1000);

Dataset<Row> ss = ora_df_reader
            .option("inferSchema", true)
            .option("schema","schema1")
            .option("numPartitions", 20);
             .option("partitionColumn", "INSERTION_DATE");
            .option("lowerBound", "2002-03-31" )
            .option("upperBound", "2019-05-01")
            .option("dateFormat", "yyyy-MM-dd" )// Tried all "yyyy-mm-dd" ,"yyyy-MM-dd" "YYYY-MM-DD" "DD-MMM-YY" "dd-MMM-yy"
            .option("dbtable", "select * from schema1.modal_vals") 
            .load();

我尝试不同方法时出现错误：

尝试过5ht方式：

.option("lowerBound", "2002-03-31 00:00:00"); 
.option("upperBound", "2019-05-01 23:59:59");
.option("timestampFormat", "yyyy-mm-dd hh:mm:ss");
.option("numPartitions", 240);

但给出错误：

java.lang.IllegalArgumentException: Timestamp format must be yyyy-mm-dd hh:mm:ss[.fffffffff]
    at java.sql.Timestamp.valueOf(Timestamp.java:204)
    at org.apache.spark.sql.execution.datasources.jdbc.JDBCRelation$.toInternalBoundValue(JDBCRelation.scala:179)

尝试了第二种方式：

.option("lowerBound","2002-03-31");
.option("upperBound", "2019-05-01");
.option("dateFormat", "yyyy-mm-dd");

我收到错误：

ORA-01861: 文字与格式字符串不匹配

如何传递“lower/upperBound”的日期？

错误：

java.sql.SQLException: ORA-12801: error signaled in parallel query server P001(2)
ORA-01861: literal does not match format string
ORA-02063: preceding 2 lines from CAPDBPROD

    at oracle.jdbc.driver.T4CTTIoer.processError(T4CTTIoer.java:445)
    at oracle.jdbc.driver.T4CTTIoer.processError(T4CTTIoer.java:396)
    at oracle.jdbc.driver.T4C8Oall.processError(T4C8Oall.java:879)
    at oracle.jdbc.driver.T4CTTIfun.receive(T4CTTIfun.java:450)
    at oracle.jdbc.driver.T4CTTIfun.doRPC(T4CTTIfun.java:192)
    at oracle.jdbc.driver.T4C8Oall.doOALL(T4C8Oall.java:531)
    at oracle.jdbc.driver.T4CPreparedStatement.doOall8(T4CPreparedStatement.java:207)
    at oracle.jdbc.driver.T4CPreparedStatement.executeForDescribe(T4CPreparedStatement.java:884)
    at oracle.jdbc.driver.OracleStatement.executeMaybeDescribe(OracleStatement.java:1167)
    at oracle.jdbc.driver.OracleStatement.doExecuteWithTimeout(OracleStatement.java:1289)
    at oracle.jdbc.driver.OraclePreparedStatement.executeInternal(OraclePreparedStatement.java:3584)
    at oracle.jdbc.driver.OraclePreparedStatement.executeQuery(OraclePreparedStatement.java:3628)
    at oracle.jdbc.driver.OraclePreparedStatementWrapper.executeQuery(OraclePreparedStatementWrapper.java:1493)

根据此修复，我们可以使用日期/时间戳作为分区列：https://issues.apache.org/jira/browse/SPARK-22814

尝试了第二种方式：

.option("lowerBound","03/31/2002 00:00:00"); 
.option("upperBound", "05/01/2019 23:59:59");
.option("dateFormat", "mm/dd/yyyy hh:mm:ss");



java.lang.IllegalArgumentException
    at java.sql.Date.valueOf(Date.java:143)
at org.apache.spark.sql.execution.datasources.jdbc.JDBCRelation$.toInternalBoundValue(JDBCRelation.scala:178)

尝试了第三种方法：

.option("lowerBound","03/31/2002"); 
.option("upperBound", "05/01/2019");
.option("dateFormat", "mm/dd/yyyy");


java.lang.IllegalArgumentException
    at java.sql.Date.valueOf(Date.java:143)
    at org.apache.spark.sql.execution.datasources.jdbc.JDBCRelation$.toInternalBoundValue(JDBCRelation.scala:178)

尝试了第四种方法：

option("lowerBound", "31.03.2002 00:00:00" );
option("upperBound", "01.05.2019 23:59:59");
option("dateFormat", "DD.MM.YYYY HH24:MI:SS")

java.lang.IllegalArgumentException 在 java.sql.Date.valueOf(Date.java:143) 在 org.apache.spark.sql.execution.datasources.jdbc.JDBCRelation$.toInternalBoundValue(JDBCRelation.scala:178)

apache-spark oracle11g apache-spark-sql oracle10g oracle11gr2
3个回答
3
投票

如果我们在下面添加它将解决问题。

option("sessionInitStatement", "ALTER SESSION SET NLS_DATE_FORMAT = 'YYYY-MM-DD'");
1
投票

当使用时间戳列作为

partitionColumn
时，下面的代码对我有用：

spark.read.format("jdbc")\
    .option("url",db_url)\
    .option("dbtable",pushdownquery)\
    .option("user",db_username)\
    .option("password",db_password)\
    .option('partitionColumn',"cdc_time")\
    .option('lowerBound','2023-01-27 00:00:00')\
    .option('upperBound','2023-01-28 00:00:00')\
    .option('numPartitions',5)\
    .option("sessionInitStatement", "ALTER SESSION SET NLS_DATE_FORMAT = 'YYYY-MM-DD HH24:MI:SS'")\
    .option('fetchSize',2000).load()

如果您想使用 date 而不是 timestamp，请添加以下选项：

.option("oracle.jdbc.mapDateToTimestamp", "false")
0
投票

在 sprk 3.2.x 中使用 Databricks 尝试上述建议并获得无效月份，有人可以尝试一下吗？

最新问题
© www.soinside.com 2019 - 2024. All rights reserved.