实践数据湖iceberg 第二十课 flink + iceberg CDC场景(版本问题，测试失败)

系列文章目录

实践数据湖iceberg 第一课 入门
实践数据湖iceberg 第二课 iceberg基于hadoop的底层数据格式
实践数据湖iceberg 第三课 在sqlclient中，以sql方式从kafka读数据到iceberg
实践数据湖iceberg 第四课 在sqlclient中，以sql方式从kafka读数据到iceberg（升级版本到flink1.12.7）
实践数据湖iceberg 第五课 hive catalog特点
实践数据湖iceberg 第六课 从kafka写入到iceberg失败问题 解决
实践数据湖iceberg 第七课 实时写入到iceberg
实践数据湖iceberg 第八课 hive与iceberg集成
实践数据湖iceberg 第九课 合并小文件
实践数据湖iceberg 第十课 快照删除
实践数据湖iceberg 第十一课 测试分区表完整流程(造数、建表、合并、删快照)
实践数据湖iceberg 第十二课 catalog是什么
实践数据湖iceberg 第十三课 metadata比数据文件大很多倍的问题
实践数据湖iceberg 第十四课 元数据合并(解决元数据随时间增加而元数据膨胀的问题)
实践数据湖iceberg 第十五课 spark安装与集成iceberg(jersey包冲突)
实践数据湖iceberg 第十六课 通过spark3打开iceberg的认知之门
实践数据湖iceberg 第十七课 hadoop2.7,spark3 on yarn运行iceberg配置
实践数据湖iceberg 第十八课 多种客户端与iceberg交互启动命令(常用命令)
实践数据湖iceberg 第十九课 flink count iceberg，无结果问题
实践数据湖iceberg 第二十课 flink + iceberg CDC场景(版本问题，测试失败)
实践数据湖iceberg 第二十一课 flink1.13.5 + iceberg0.131 CDC(测试成功)

---

---

概要

测试flink cdc, 以及数据变化时update数据是如何落地

flink1.14.3
iceberg0.13.0
cdc: 2.2

设计测试场景：

---

1. mysql数据准备

1.1 准备数据(初始化)

create database xxzh_stock character set utf8;

CREATE TABLE `stock_basic` (
  `i` bigint(20) DEFAULT NULL,
  `ts_code` varchar(10),
  `symbol` varchar(10),
  `name` varchar(10),
  `area` varchar(20),
  `industry` varchar(20),
  `list_date` varchar(10),
  `actural_controller` varchar(100) DEFAULT NULL,
  KEY `ix_stock_basic_index` (`i`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;



-- ----------------------------
-- Records of stock_basic
-- ----------------------------
INSERT INTO `stock_basic` VALUES ('0', '000001.SZ', '000001', '平安银行', '深圳', '银行', '19910403', null);
INSERT INTO `stock_basic` VALUES ('1', '000002.SZ', '000002', '万科A', '深圳', '全国地产', '19910129', null);
INSERT INTO `stock_basic` VALUES ('2', '000004.SZ', '000004', '国华网安', '深圳', '软件服务', '19910114', '李映彤');

1.2 开启mysql binlog

修改mysql配置，/etc/my.cnf，开启binlog, 并记录binlog-do-db=xxzh_stock 这个库的变化
文件末尾，增加如下配置

server-id=1
log-bin=mysql-bin
binlog_format=row
binlog-do-db=xxzh_stock

重启mysql，让配置生效

[root@hadoop103 conf]# service mysqld restart

2 flink表定义

2.1 启动flink-sql命令

带上icerbg,kafka,hive,cdc的包

[root@hadoop101 ~]# sql-client.sh embedded -j /opt/software/iceberg0.13/iceberg-flink-runtime-1.14-0.13.0.jar -j /opt/software/iceberg0.13/flink-sql-connector-hive-2.3.6_2.12-1.14.3.jar -j /opt/software/flink-sql-connector-kafka_2.12-1.14.3.jar -j /opt/software/flink-connector-mysql-cdc-1.4.0.jar shell

cdc的下载地址： https://repo.maven.apache.org/maven2/com/alibaba/ververica/flink-connector-mysql-cdc/1.4.0/flink-connector-mysql-cdc-1.4.0.jar

2.2 准备mysql的表source表

CREATE TABLE stock_basic_source(
  `i`  INT NOT NULL,
  `ts_code`   INT NOT NULL,
  `symbol`   CHAR(10) NOT NULL,
  `name` char(10) NOT NULL,
  `area`   CHAR(20) NOT NULL,
  `industry`   CHAR(20) NOT NULL,
  `list_date`   CHAR(10) NOT NULL,
  `actural_controller`   CHAR(100) NOT NULL,
  PRIMARY KEY(i) NOT ENFORCED
) WITH (
  'connector' = 'mysql-cdc',
  'hostname' = 'hadoop103',
  'port' = '3306',
  'username' = 'root',
  'password' = '123456',
  'database-name' = 'xxzh_stock',
  'table-name' = 'stock_basic'
);

PRIMARY KEY(i) NOT ENFORCED需要加上主键，否则会报

Flink SQL> select * from stock_basic_source;
[ERROR] Could not execute SQL statement. Reason:
org.apache.flink.table.api.ValidationException: The primary key is necessary when enable 'Key: 'scan.incremental.snapshot.enabled' , default: true (fallback keys: [])' to 'true'

2.3 准备iceberg的sink表

CREATE CATALOG hive_catalog6 WITH (
  'type'='iceberg',
  'catalog-type'='hive',
  'uri'='thrift://hadoop101:9083',
  'clients'='5',
  'property-version'='1',
  'warehouse'='hdfs://user/hive/warehouse/hive_catalog6'
);
use catalog hive_catalog6;

CREATE DATABASE xxzh_stock_mysql_db;
USE xxzh_stock_mysql_db;

CREATE TABLE stock_basic_iceberg_sink(
  `i`  INT NOT NULL,
  `ts_code`   INT NOT NULL,
  `symbol`   CHAR(10) NOT NULL,
  `name` char(10) NOT NULL,
  `area`   CHAR(20) NOT NULL,
  `industry`   CHAR(20) NOT NULL,
  `list_date`   CHAR(10) NOT NULL,
  `actural_controller`   CHAR(100) NOT NULL,
   PRIMARY KEY(i) NOT ENFORCED
);

3.通过flink从mysq写入iceberg

3.1 数据由mysql写入到iceberg

Flink SQL> insert into hive_catalog6.xxzh_stock_mysql_db.stock_basic_iceberg_sink select * from stock_basic_source;
[INFO] Submitting SQL update statement to the cluster...
[INFO] SQL update statement has been successfully submitted to the cluster:
Job ID: c83f55f408926ad987375f4fa3bf7df4

4. catalog路径说明

catalog切换回default_catalog才能查soure表，或者带default_catalog的绝对路径

4.1 切换catalog方式

Flink SQL> select * from stock_basic_source;
[ERROR] Could not execute SQL statement. Reason:
org.apache.calcite.sql.validate.SqlValidatorException: Object 'stock_basic_source' not found

Flink SQL> show current catalog;
+----------------------+
| current catalog name |
+----------------------+
|        hive_catalog6 |
+----------------------+
1 row in set

Flink SQL> show catalogs;
+-----------------+
|    catalog name |
+-----------------+
| default_catalog |
|   hive_catalog6 |
+-----------------+
2 rows in set

Flink SQL> use catalog default_catalog;
[INFO] Execute statement succeed.

Flink SQL> select * from stock_basic_source;

4.2 绝对路径方式

参考3.2

5 报错处理

5.1 cdc包的问题

查询

解决方法：准备这个包 flink-connector-mysql-cdc-1.4.0.jar

Flink SQL> select * from stock_basic_source;
[ERROR] Could not execute SQL statement. Reason:
org.apache.flink.table.api.ValidationException: Could not find any factory for identifier 'mysql-cdc' that implements 'org.apache.flink.table.factories.DynamicTableFactory' in the classpath.

Available factory identifiers are:

blackhole
datagen
filesystem
iceberg
kafka
print
upsert-kafka

5.2 java.lang.NoClassDefFoundError: org/apache/flink/shaded/guava18/com/google/common/util/concurrent/ThreadFactoryBuilder

2022-02-17 16:24:47,934 INFO  org.apache.flink.runtime.state.StateBackendLoader            [] - State backend loader loads the state backend as HashMapStateBackend
2022-02-17 16:24:47,934 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using job/cluster config to configure application-defined checkpoint storage: org.apache.flink.runtime.state.storage.FileSystemCheckpointStorage@2772b695
2022-02-17 16:24:47,936 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using job/cluster config to configure application-defined state backend: org.apache.flink.runtime.state.hashmap.HashMapStateBackend@16ba5737
2022-02-17 16:24:47,936 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using application-defined state backend: org.apache.flink.runtime.state.hashmap.HashMapStateBackend@4ae44fbc
2022-02-17 16:24:47,936 INFO  org.apache.flink.runtime.state.StateBackendLoader            [] - State backend loader loads the state backend as HashMapStateBackend
2022-02-17 16:24:47,936 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using job/cluster config to configure application-defined checkpoint storage: org.apache.flink.runtime.state.storage.FileSystemCheckpointStorage@276604d1
2022-02-17 16:24:47,936 INFO  org.apache.flink.runtime.taskmanager.Task                    [] - SinkMaterializer -> IcebergStreamWriter (1/1)#3 (f31a1159e5464010d0ea8c9325e3db32) switched from DEPLOYING to INITIALIZING.
2022-02-17 16:24:47,941 INFO  org.apache.flink.runtime.taskmanager.Task                    [] - Source: TableSourceScan(table=[[default_catalog, default_database, stock_basic_source]], fields=[i, ts_code, symbol, name, area, industry, list_date, actural_controller]) -> NotNullEnforcer(fields=[i, ts_code, symbol, name, area, industry, list_date, actural_controller]) (1/1)#3 (3b4e39607a907745d88639c7300b088a) switched from DEPLOYING to INITIALIZING.
2022-02-17 16:24:47,942 WARN  org.apache.flink.metrics.MetricGroup                         [] - The operator name NotNullEnforcer(fields=[i, ts_code, symbol, name, area, industry, list_date, actural_controller]) exceeded the 80 characters length limit and was truncated.
2022-02-17 16:24:47,943 INFO  org.apache.flink.runtime.taskexecutor.TaskExecutor           [] - Received task IcebergFilesCommitter -> Sink: IcebergSink hive_catalog6.xxzh_stock_mysql_db.stock_basic_iceberg_sink (1/1)#3 (9dc85e56ba89021f64f46a5c55b86e9c), deploy into slot with allocation id 6fa395ed1d6e1678faaa162efd8ca0ff.
2022-02-17 16:24:47,946 WARN  org.apache.flink.metrics.MetricGroup                         [] - The operator name Source: TableSourceScan(table=[[default_catalog, default_database, stock_basic_source]], fields=[i, ts_code, symbol, name, area, industry, list_date, actural_controller]) exceeded the 80 characters length limit and was truncated.
2022-02-17 16:24:47,946 INFO  com.alibaba.ververica.cdc.debezium.DebeziumSourceFunction    [] - Consumer subtask 0 has no restore state.
2022-02-17 16:24:47,947 INFO  org.apache.flink.runtime.taskmanager.Task                    [] - IcebergFilesCommitter -> Sink: IcebergSink hive_catalog6.xxzh_stock_mysql_db.stock_basic_iceberg_sink (1/1)#3 (9dc85e56ba89021f64f46a5c55b86e9c) switched from CREATED to DEPLOYING.
2022-02-17 16:24:47,947 INFO  org.apache.flink.runtime.taskmanager.Task                    [] - Loading JAR files for task IcebergFilesCommitter -> Sink: IcebergSink hive_catalog6.xxzh_stock_mysql_db.stock_basic_iceberg_sink (1/1)#3 (9dc85e56ba89021f64f46a5c55b86e9c) [DEPLOYING].
2022-02-17 16:24:47,947 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using job/cluster config to configure application-defined state backend: org.apache.flink.runtime.state.hashmap.HashMapStateBackend@5d60868d
2022-02-17 16:24:47,947 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using application-defined state backend: org.apache.flink.runtime.state.hashmap.HashMapStateBackend@595f09f9
2022-02-17 16:24:47,948 INFO  org.apache.flink.runtime.state.StateBackendLoader            [] - State backend loader loads the state backend as HashMapStateBackend
2022-02-17 16:24:47,948 INFO  org.apache.flink.streaming.runtime.tasks.StreamTask          [] - Using job/cluster config to configure application-defined checkpoint storage: org.apache.flink.runtime.state.storage.FileSystemCheckpointStorage@7406169e
2022-02-17 16:24:47,948 INFO  org.apache.flink.runtime.taskmanager.Task                    [] - IcebergFilesCommitter -> Sink: IcebergSink hive_catalog6.xxzh_stock_mysql_db.stock_basic_iceberg_sink (1/1)#3 (9dc85e56ba89021f64f46a5c55b86e9c) switched from DEPLOYING to INITIALIZING.
2022-02-17 16:24:47,955 WARN  org.apache.flink.runtime.taskmanager.Task                    [] - Source: TableSourceScan(table=[[default_catalog, default_database, stock_basic_source]], fields=[i, ts_code, symbol, name, area, industry, list_date, actural_controller]) -> NotNullEnforcer(fields=[i, ts_code, symbol, name, area, industry, list_date, actural_controller]) (1/1)#3 (3b4e39607a907745d88639c7300b088a) switched from INITIALIZING to FAILED with failure cause: java.lang.NoClassDefFoundError: org/apache/flink/shaded/guava18/com/google/common/util/concurrent/ThreadFactoryBuilder
	at com.alibaba.ververica.cdc.debezium.DebeziumSourceFunction.open(DebeziumSourceFunction.java:166)
	at org.apache.flink.api.common.functions.util.FunctionUtils.openFunction(FunctionUtils.java:34)
	at org.apache.flink.streaming.api.operators.AbstractUdfStreamOperator.open(AbstractUdfStreamOperator.java:100)
	at org.apache.flink.streaming.runtime.tasks.RegularOperatorChain.initializeStateAndOpenOperators(RegularOperatorChain.java:110)
	at org.apache.flink.streaming.runtime.tasks.StreamTask.restoreGates(StreamTask.java:711)
	at org.apache.flink.streaming.runtime.tasks.StreamTaskActionExecutor$SynchronizedStreamTaskActionExecutor.call(StreamTaskActionExecutor.java:100)
	at org.apache.flink.streaming.runtime.tasks.StreamTask.restoreInternal(StreamTask.java:687)
	at org.apache.flink.streaming.runtime.tasks.StreamTask.restore(StreamTask.java:654)
	at org.apache.flink.runtime.taskmanager.Task.runWithSystemExitMonitoring(Task.java:958)
	at org.apache.flink.runtime.taskmanager.Task.restoreAndInvoke(Task.java:927)
	at org.apache.flink.runtime.taskmanager.Task.doRun(Task.java:766)
	at org.apache.flink.runtime.taskmanager.Task.run(Task.java:575)
	at java.lang.Thread.run(Thread.java:748)

问题分析：
cdc目前支持到flink1.13.3,具体看github pom, https://github.com/ververica/flink-cdc-connectors/blob/master/pom.xml

解决方法：
降级或者编译一个cdc，参考 http://betheme.net/news/txtlist_i66236v.html

碰到这种问题，很难讲心情好。。。

下载源码 ： https://github.com/ververica/flink-cdc-connectors

尝试编译：发现1.14.3的blink包是没有的，好吧，还是放弃好了。blink-planner没有，未来吃太多坑，找社区的人问(github上有个cdc的社区服务钉钉群)，也不建议使用flink1.14。

降级！
最新版flink1.13的子版本是flink1.13.5,cdc的flink1.13.3, 重新编译一个cdc
cdc的pom.xml中修改flink和scala版本。/

[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary for flink-cdc-connectors 2.2-SNAPSHOT:
[INFO] 
[INFO] flink-cdc-connectors ............................... SUCCESS [ 11.822 s]
[INFO] flink-connector-debezium ........................... SUCCESS [  6.094 s]
[INFO] flink-connector-test-util .......................... SUCCESS [  0.812 s]
[INFO] flink-connector-mysql-cdc .......................... SUCCESS [ 11.715 s]
[INFO] flink-connector-postgres-cdc ....................... SUCCESS [  1.001 s]
[INFO] flink-connector-oracle-cdc ......................... SUCCESS [  1.162 s]
[INFO] flink-connector-mongodb-cdc ........................ SUCCESS [  1.119 s]
[INFO] flink-connector-sqlserver-cdc ...................... SUCCESS [  0.818 s]
[INFO] flink-sql-connector-mysql-cdc ...................... SUCCESS [  6.187 s]
[INFO] flink-sql-connector-postgres-cdc ................... SUCCESS [  4.314 s]
[INFO] flink-sql-connector-mongodb-cdc .................... SUCCESS [  4.149 s]
[INFO] flink-sql-connector-oracle-cdc ..................... SUCCESS [  6.307 s]
[INFO] flink-sql-connector-sqlserver-cdc .................. SUCCESS [  3.716 s]
[INFO] flink-format-changelog-json ........................ SUCCESS [  0.321 s]
[INFO] flink-cdc-e2e-tests ................................ SUCCESS [  0.906 s]
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  01:00 min
[INFO] Finished at: 2022-02-17T21:00:10+08:00
[INFO] ------------------------------------------------------------------------

---

总结

一开始，没有把版本全面考虑，现在又要回滚版本，过程痛苦