SparkSQL JDBC 查询Oracle、MySQL 转化为Hudi表

前言

本文总结如何只用SQL迁移关系型数据库中的表转化为Hudi表，这样的意义在于方便项目上对Spark、Hudi等不熟悉的人员运维。

Spark Thrift Server

首先将Orace的驱动拷贝至Spark jars目录下

启动Spark Thrift Server 扩展支持Hudi

1
/usr/hdp/3.1.0.0-78/spark2/bin/spark-submit --master yarn --deploy-mode client --executor-memory 2G --num-executors 6 --executor-cores 2 --driver-memory 4G --driver-cores 2 --class org.apache.spark.sql.hive.thriftserver.HiveThriftServer2 --name Thrift-20003 --conf 'spark.serializer=org.apache.spark.serializer.KryoSerializer'  --conf 'spark.sql.extensions=org.apache.spark.sql.hudi.HoodieSparkSessionExtension' --principal spark/indata-192-168-44-128.indata.com@INDATA.COM --keytab /etc/security/keytabs/spark.service.keytab --hiveconf hive.server2.thrift.http.port=20003 >~/server.log 2>&1 &

Beeline连接

1
/usr/hdp/3.1.0.0-78/spark2/bin/beeline -u "jdbc:hive2://192.168.44.128:20003/hudi;principal=HTTP/indata-192-168-44-128.indata.com@INDATA.COM?hive.server2.transport.mode=http;hive.server2.thrift.http.path=cliservice"
more >>

展开全文 >>

Spark Thrift Server HA 解决方案

前言

本文为组内同事整理,这里稍作改动记录。本文主要修改Spark源码，实现Spark Spark Thrift Server 注册到到ZK，通过ZK连接实现负载均衡；另外可以通过使用Kyuubi实现HA，这里不做详细描述

背景

  • Spark ThriftServer 不支持Zookeeper连接，不能实现负载均衡。
  • 解决方案:修改Spark ThriftServer源码，使其支持Zookeeper连接
  • 版本信息：Spark3.1.2 ；Hive2.3.7

Spark ThriftServer 启动命令

1
nohup spark-submit --master yarn --deploy-mode client --num-executors 2 --executor-cores 1 --executor-memory 1G --class org.apache.spark.sql.hive.thriftserver.HiveThriftServer2 --name SparkThriftServer_1 spark-internal --hiveconf hive.server2.thrift.http.port=20003 >>  /var/log/spark2/20003.log 2>&1 < /dev/null &

Spark 源码debug

1
./bin/spark-submit --master yarn --deploy-mode client --num-executors 2 --executor-cores 1 --executor-memory 1G --class org.apache.spark.sql.hive.thriftserver.HiveThriftServer2 --name SparkThriftServer_1 spark-internal --hiveconf hive.server2.thrift.http.port=20003  --driver-java-options "-Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=y,address=5005"

Spark 连接

单节点

1
!connect jdbc:hive2://indata-1192-168-44-128:20003/default;principal=HTTP/indata-1192-168-44-128@INDATA.COM?hive.server2.transport.mode=http;hive.server2.thrift.http.path=cliservice;

Zookeeper

1
!connect jdbc:hive2://192-168-44-129.indata.com:2181,indata-1192-168-44-128:2181,indata-192-168-44-130.indata.com:2181/default;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2_zk
more >>

展开全文 >>

Presto查询Hudi异常解决

前言

记录一个Presto查询Hudi的异常的解决办法，本人目前对Presto还不是很熟，只是在使用过程中遇到了问题，记录一下异常解决方法以及过程

异常

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
2021-12-22T17:29:55.440+0800    ERROR   SplitRunner-101-126     com.facebook.presto.execution.executor.TaskExecutor     Error processing Split 20211222_092954_00047_8xk77.1.0.0-0 {path=hdfs://cluster1/warehouse/tablespace/managed/hive/test_dkl.db/test_hudi_1/069eddc2-f3bf-4efc-911b-3cb3aa523a8e-0_0-0-0_20211222172800.parquet, start=0, length=435116, fileSize=435116, fileModifiedTime=1640165281695, hosts=[], database=test_dkl, table=test_hudi_1, nodeSelectionStrategy=NO_PREFERENCE, partitionName=<UNPARTITIONED>, s3SelectPushdownEnabled=false, cacheQuotaRequirement=CacheQuotaRequirement{cacheQuotaScope=GLOBAL, quota=Optional.empty}} (start = 9.271133085962032E9, wall = 49 ms, cpu = 0 ms, wait = 0 ms, calls = 1)
org.apache.parquet.io.ParquetDecodingException: Can not read value at 0 in block -1 in file hdfs://cluster1/warehouse/tablespace/managed/hive/test_dkl.db/test_hudi_1/069eddc2-f3bf-4efc-911b-3cb3aa523a8e-0_0-0-0_20211222172800.parquet
        at org.apache.parquet.hadoop.InternalParquetRecordReader.nextKeyValue(InternalParquetRecordReader.java:251)
        at org.apache.parquet.hadoop.ParquetRecordReader.nextKeyValue(ParquetRecordReader.java:207)
        at org.apache.hadoop.hive.ql.io.parquet.read.ParquetRecordReaderWrapper.<init>(ParquetRecordReaderWrapper.java:98)
        at org.apache.hadoop.hive.ql.io.parquet.read.ParquetRecordReaderWrapper.<init>(ParquetRecordReaderWrapper.java:60)
        at org.apache.hadoop.hive.ql.io.parquet.MapredParquetInputFormat.getRecordReader(MapredParquetInputFormat.java:92)
        at org.apache.hudi.hadoop.HoodieParquetInputFormat.getRecordReader(HoodieParquetInputFormat.java:216)
        at com.facebook.presto.hive.HiveUtil.createRecordReader(HiveUtil.java:266)
        at com.facebook.presto.hive.GenericHiveRecordCursorProvider.lambda$createRecordCursor$0(GenericHiveRecordCursorProvider.java:74)
        at com.facebook.presto.hive.authentication.UserGroupInformationUtils.lambda$executeActionInDoAs$0(UserGroupInformationUtils.java:29)
        at java.security.AccessController.doPrivileged(Native Method)
        at javax.security.auth.Subject.doAs(Subject.java:360)
        at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1726)
        at com.facebook.presto.hive.authentication.UserGroupInformationUtils.executeActionInDoAs(UserGroupInformationUtils.java:27)
        at com.facebook.presto.hive.authentication.DirectHdfsAuthentication.doAs(DirectHdfsAuthentication.java:38)
        at com.facebook.presto.hive.HdfsEnvironment.doAs(HdfsEnvironment.java:81)
        at com.facebook.presto.hive.GenericHiveRecordCursorProvider.createRecordCursor(GenericHiveRecordCursorProvider.java:73)
        at com.facebook.presto.hive.HivePageSourceProvider.createHivePageSource(HivePageSourceProvider.java:478)
        at com.facebook.presto.hive.HivePageSourceProvider.createPageSource(HivePageSourceProvider.java:184)
        at com.facebook.presto.spi.connector.classloader.ClassLoaderSafeConnectorPageSourceProvider.createPageSource(ClassLoaderSafeConnectorPageSourceProvider.java:63)
        at com.facebook.presto.split.PageSourceManager.createPageSource(PageSourceManager.java:80)
        at com.facebook.presto.operator.TableScanOperator.getOutput(TableScanOperator.java:249)
        at com.facebook.presto.operator.Driver.processInternal(Driver.java:424)
        at com.facebook.presto.operator.Driver.lambda$processFor$9(Driver.java:307)
        at com.facebook.presto.operator.Driver.tryWithLock(Driver.java:728)
        at com.facebook.presto.operator.Driver.processFor(Driver.java:300)
        at com.facebook.presto.execution.SqlTaskExecution$DriverSplitRunner.processFor(SqlTaskExecution.java:1077)
        at com.facebook.presto.execution.executor.PrioritizedSplitRunner.process(PrioritizedSplitRunner.java:162)
        at com.facebook.presto.execution.executor.TaskExecutor$TaskRunner.run(TaskExecutor.java:599)
        at com.facebook.presto.$gen.Presto_0_265_1_ad1fce6____20211222_010139_1.run(Unknown Source)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
        at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.UnsupportedOperationException: readDirect unsupported in RemoteBlockReader
        at org.apache.hadoop.hdfs.RemoteBlockReader.read(RemoteBlockReader.java:492)
        at org.apache.hadoop.hdfs.DFSInputStream$ByteBufferStrategy.doRead(DFSInputStream.java:789)
        at org.apache.hadoop.hdfs.DFSInputStream.readBuffer(DFSInputStream.java:823)
        at org.apache.hadoop.hdfs.DFSInputStream.readWithStrategy(DFSInputStream.java:883)
        at org.apache.hadoop.hdfs.DFSInputStream.read(DFSInputStream.java:938)
        at org.apache.hadoop.fs.FSDataInputStream.read(FSDataInputStream.java:143)
        at org.apache.parquet.hadoop.util.H2SeekableInputStream$H2Reader.read(H2SeekableInputStream.java:81)
        at org.apache.parquet.hadoop.util.H2SeekableInputStream.readFully(H2SeekableInputStream.java:90)
        at org.apache.parquet.hadoop.util.H2SeekableInputStream.readFully(H2SeekableInputStream.java:75)
        at org.apache.parquet.hadoop.ParquetFileReader$ConsecutiveChunkList.readAll(ParquetFileReader.java:1174)
        at org.apache.parquet.hadoop.ParquetFileReader.readNextRowGroup(ParquetFileReader.java:805)
        at org.apache.parquet.hadoop.InternalParquetRecordReader.checkRead(InternalParquetRecordReader.java:127)
        at org.apache.parquet.hadoop.InternalParquetRecordReader.nextKeyValue(InternalParquetRecordReader.java:222)
more >>

展开全文 >>

利用Submarin集成Spark-Ranger

前言

本文总结如何利用Submarin集成Spark-Ranger，通过ranger控制Spark SQL的权限

前提

已经安装了Spark、Hive、kerberos、Ranger,并且Hive已经集成了Ranger,本文环境基于Ambari

submarine-spark-security 插件打包

官网文档https://submarine.apache.org/docs/userDocs/submarine-security/spark-security/README (0.6.0版本)

1
git clone https://github.com/apache/submarine

当前master分支（0.7.0）已经没有submarine-spark-security模块了，需要切换到tag:release-0.6.0-RC0

然后利用mvn命令打包

1
mvn clean package -Dmaven.javadoc.skip=true -DskipTests -pl :submarine-spark-security -Pspark-2.4 -Pranger-1.2

打的submarine-spark-security-0.6.0.jar在目录submarine\submarine-security\spark-security\target

然后将submarine-spark-security-0.6.0.jar上传到$SPARK_HOME/jars

more >>

展开全文 >>

Hudi preCombinedField 总结(二)-源码分析

前言

在上一篇博客Hudi preCombinedField 总结中已经对preCombinedField进行总结过一次了，由于当时对源码理解还不够深入，导致分析的不全面，现在对源码有了进一步的理解，所以再进行总结补充一下。

历史比较值

上面总结中：

DF：无论新记录的ts值是否大于历史记录的ts值，都会覆盖写，直接更新。
SQL:写数据时，ts值大于等于历史ts值，才会更新，小于历史值则不更新。

这里解释一下原因，首先Spark SQL PAYLOAD_CLASS_NAME 默认值为ExpressionPayload，而ExpressionPayload继承了DefaultHoodieRecordPayload

1
2
3
class ExpressionPayload(record: GenericRecord,
                        orderingVal: Comparable[_])
  extends DefaultHoodieRecordPayload(record, orderingVal) {

DefaultHoodieRecordPayload 里的needUpdatingPersistedRecord实现了历史值进行比较，具体实现，后面会进行分析

而 Spark DF在hudi0.9.0版本 PAYLOAD_CLASS_NAME的默认值为OverwriteWithLatestAvroPayload，它是DefaultHoodieRecordPayload的父类并没有实现和历史值进行比较

more >>

展开全文 >>

tag:

    缺失模块。
    1、请确保node版本大于6.2
    2、在博客根目录（注意不是yilia根目录）执行以下命令：
    npm i hexo-generator-json-content --save

    3、在根目录_config.yml里添加配置： 

      jsonContent:
    meta: false
    pages: false
    posts:
      title: true
      date: true
      path: true
      text: false
      raw: false
      content: false
      slug: false
      updated: false
      comments: false
      link: false
      permalink: false
      excerpt: false
      categories: false
      tags: true

2016年本科毕业于合肥工业大学<br><br>信息与计算科学专业<br><br>目前在济南一家公司从事大数据的工作~~