[
https://issues.apache.org/jira/browse/FLINK-17459?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17096029#comment-17096029
]
ranqiqiang edited comment on FLINK-17459 at 4/30/20, 2:53 AM:
--------------------------------------------------------------
1.JDBCAppendTableSink will be abandoned ?
2. I don't know JDBCUpsertTableSink#Builder why not "isAppendOnly" and
"keyFields"
not do like this:
{code:java}
//代码占位符
JDBCUpsertTableSink sink = JDBCUpsertTableSink.builder()
.setOptions(options)
.setTableSchema(schema)
.setFlushIntervalMills(3000)
.setIsAppendOnly(true) // not support
.build();
// have to do like
sink.setIsAppendOnly(true);
// or
sink.setKeyFields(new String[]{"id"});
{code}
was (Author: michael ran):
1.JDBCAppendTableSink will be abandoned ?
> JDBCAppendTableSink not support flush by flushIntervalMills
> --------------------------------------------------------------
>
> Key: FLINK-17459
> URL: https://issues.apache.org/jira/browse/FLINK-17459
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / JDBC
> Affects Versions: 1.10.0
> Reporter: ranqiqiang
> Priority: Major
>
> {{JDBCAppendTableSink just support append by
> "JDBCAppendTableSinkBuilder#batchSize",}}{{not support like
> "JDBCUpsertTableSink#flushIntervalMills"}}
>
> {{If batchSize=5000 , my data rows=5000*N+1 ,then last one record could not
> be append !!}}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
