Hi Yan, I’d like to look into this. Can you share more about your queries and the full stack trace?
Thank, Xingcan > On 8 Mar 2018, at 11:28 AM, Yan Zhou [FDS Science] <yz...@coupang.com > <mailto:yz...@coupang.com>> wrote: > > Hi experts, > I am using flink table api to join two tables, which are datastream > underneath. However, I got an assertion error of "java.lang.AssertionError: > mismatched type $1 TIMESTAMP(3)" on rowtime column. Below is more details: > > There in only one kafka data source, which is then converted to Table and > registered. One existed column is set as rowtime(event time) attribute. Two > over-window aggregation queries are run against the table and two tables are > created as results. Everything works great so far. > However when timed-window joining two result tables with inherented rowtime, > calcite throw the "java.lang.AssertionError: mismatched type $1 TIMESTAMP(3)" > AssertionError. Can someone let me know what is the possible cause? F.Y.I., I > rename the rowtime column for one of the result table. > > DataStream<MyObject> dataStream = env.addSource(kafkaConsumer); > Table table = tableEnv.fromDataStream(dataStream, "col1", "col2", ...); > tableEnv.registerTable(tableName, table); > Table left = tableEnv.sqlQuery("select id, eventTime,count (*) over ... from > ..."); > Table right = tableEnv.sqlQuery("select id as r_id, eventTime as > r_event_time, count (*) over ... from ..."); > left.join(right).where("id = r_id && eventTime === r_event_time) > .addSink(...); // here calcite throw exception: java.lang.AssertionError: > mismatched type $1 TIMESTAMP(3) > > source table > |-- id: Long > |-- eventTime: TimeIndicatorTypeInfo(rowtime) > |-- ... > |-- ... > > result_1 table > |-- id: Long > |-- eventTime: TimeIndicatorTypeInfo(rowtime) > |-- ... > |-- ... > > result_2 table > |-- rid: Long > |-- r_event_time: TimeIndicatorTypeInfo(rowtime) > |-- ... > > > Best > Yan
