我一直在努力解决这个问题大约一个星期,现在试图获取一个简单的(3个字段)AVRO格式的KSQL表作为JDBC连接器接收器(mysql)的源
我收到以下错误(在INFO行之后):
[2018-12-11 18:58:50,678] INFO Setting metadata for table "DSB_ERROR_TABLE_WINDOWED" to Table{name='"DSB_ERROR_TABLE_WINDOWED"', columns=[Column{'MOD_CLASS', isPrimaryKey=false, allowsNull=true, sqlType=VARCHAR}, Column{'METHOD', isPrimaryKey=false, allowsNull=true, sqlType=VARCHAR}, Column{'COUNT', isPrimaryKey=false, allowsNull=true, sqlType=BIGINT}]} (io.confluent.connect.jdbc.util.TableDefinitions)
[2018-12-11 18:58:50,679] ERROR WorkerSinkTask{id=dev-dsb-errors-mysql-sink-0} Task threw an uncaught and unrecoverable exception. Task is being killed and will not recover until manually restarted. (org.apache.kafka.connect.runtime.WorkerSinkTask)
org.apache.kafka.connect.errors.ConnectException: No fields found using key and value schemas for table: DSB_ERROR_TABLE_WINDOWED
at io.confluent.connect.jdbc.sink.metadata.FieldsMetadata.extract(FieldsMetadata.java:127)
at io.confluent.connect.jdbc.sink.metadata.FieldsMetadata.extract(FieldsMetadata.java:64)
at io.confluent.connect.jdbc.sink.BufferedRecords.add(BufferedRecords.java:79)
at io.confluent.connect.jdbc.sink.BufferedRecords.add(BufferedRecords.java:124)
at io.confluent.connect.jdbc.sink.JdbcDbWriter.write(JdbcDbWriter.java:63)
at io.confluent.connect.jdbc.sink.JdbcSinkTask.put(JdbcSinkTask.java:75)
at org.apache.kafka.connect.runtime.WorkerSinkTask.deliverMessages(WorkerSinkTask.java:564)
at org.apache.kafka.connect.runtime.WorkerSinkTask.poll(WorkerSinkTask.java:322)
at org.apache.kafka.connect.runtime.WorkerSinkTask.iteration(WorkerSinkTask.java:225)
at org.apache.kafka.connect.runtime.WorkerSinkTask.execute(WorkerSinkTask.java:193)
at org.apache.kafka.connect.runtime.WorkerTask.doRun(WorkerTask.java:175)
at org.apache.kafka.connect.runtime.WorkerTask.run(WorkerTask.java:219)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
我可以断定,在拉出架构时,接收器正在做正确的事情(请参见上面的错误之前),并且使用正确的架构在数据库中成功创建了表:
MariaDB [dsb_errors_ksql]> describe DSB_ERROR_TABLE_WINDOWED;
+-----------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-----------+--------------+------+-----+---------+-------+
| MOD_CLASS | varchar(256) | YES | | NULL | |
| METHOD | varchar(256) | YES | | NULL | |
| COUNT | bigint(20) | YES | | NULL | |
+-----------+--------------+------+-----+---------+-------+
3 rows in set (0.01 sec)
这是KTABLE定义:
ksql> describe extended DSB_ERROR_TABLE_windowed;
Name : DSB_ERROR_TABLE_WINDOWED
Type : TABLE
Key field : KSQL_INTERNAL_COL_0|+|KSQL_INTERNAL_COL_1
Key format : STRING
Timestamp field : Not set - using <ROWTIME>
Value format : AVRO
Kafka topic : DSB_ERROR_TABLE_WINDOWED (partitions: 4, replication: 1)
Field | Type
---------------------------------------
ROWTIME | BIGINT (system)
ROWKEY | VARCHAR(STRING) (system)
MOD_CLASS | VARCHAR(STRING)
METHOD | VARCHAR(STRING)
COUNT | BIGINT
---------------------------------------
Queries that write into this TABLE
-----------------------------------
CTAS_DSB_ERROR_TABLE_WINDOWED_37 : create table DSB_ERROR_TABLE_windowed with (value_format='avro') as select mod_class, method, count(*) as count from DSB_ERROR_STREAM window session ( 60 seconds) group by mod_class, method having count(*) > 0;
在架构注册表中会为此表自动生成一个条目(但没有键条目):
{
"subject": "DSB_ERROR_TABLE_WINDOWED-value",
"version": 7,
"id": 143,
"schema": "{\"type\":\"record\",\"name\":\"KsqlDataSourceSchema\",\"namespace\":\"io.confluent.ksql.avro_schemas\",\"fields\":[{\"name\":\"MOD_CLASS\",\"type\":[\"null\",\"string\"],\"default\":null},{\"name\":\"METHOD\",\"type\":[\"null\",\"string\"],\"default\":null},{\"name\":\"COUNT\",\"type\":[\"null\",\"long\"],\"default\":null}]}"
}
这是Connect Worker的定义:
{ "name": "dev-dsb-errors-mysql-sink",
"config": {
"connector.class": "io.confluent.connect.jdbc.JdbcSinkConnector",
"tasks.max": "1",
"topics": "DSB_ERROR_TABLE_WINDOWED",
"connection.url": "jdbc:mysql://os-compute-d01.maeagle.corp:32692/dsb_errors_ksql?user=xxxxxx&password=xxxxxx",
"auto.create": "true",
"value.converter": "io.confluent.connect.avro.AvroConverter",
"value.converter.schema.registry.url": "http://kafka-d01.maeagle.corp:8081",
"key.converter": "org.apache.kafka.connect.storage.StringConverter"
}
}
我的理解(可能是错误的)是KSQL应该在Schema Registry中创建适当的AVRO Schema,而Kafka Connect应该能够正确地读取这些内容。如上所述,虽然在Mysql中生成了适当的表,但仍在正常工作,尽管我很惊讶没有创建键字段...
大多数帖子和示例都使用JSON而非AVRO,因此它们并不是特别有用。
这似乎是在读取和插入主题记录的反序列化部分...
这时我很茫然,可以使用一些指导。
我还通过github打开了类似的票证:
https://github.com/confluentinc/ksql/issues/2250
此致
-约翰
答案 0 :(得分:0)
如约翰在上面所述,主题记录中的关键字不是字符串,而是用单个Java序列化的64位整数后缀固定的字符串,表示窗口的开始时间。
Connect不附带可以处理窗口密钥格式的SMT。但是,可以写一个剥离整数并仅返回自然键。然后,您可以将其包括在类路径中并更新您的连接配置。
如果您需要数据库中的窗口开始时间,则可以更新ksqlDB查询,以将窗口开始时间作为字段包括在值中。