王太阳 created FLINK-16651:
--------------------------- Summary: flink stream SQL INNER JOIN乱序 Key: FLINK-16651 URL: https://issues.apache.org/jira/browse/FLINK-16651 Project: Flink Issue Type: Bug Reporter: 王太阳 flink1.9.1:在读取kafka的两个topic(单分区)的时候,注册为两个表并进行 INNER JOIN,数出的结果是乱序的 topic: test {code:java} a,b,c a,b,c1 a,b,c2{code} topic: test2 {code:java} a,b2,c2 a,b3,c3 a,b4,c4 a,b5,c5 a,b6,c6{code} FLINK SQL: {code:java} select * from test t inner join test2 t2 on t.a=t2.a {code} 第一次输出: {code:java} 1> (true,a,b,c,2020-03-18T09:40:11.858,a,b2,c2,2020-03-18T09:40:11.858) 1> (true,a,b,c2,2020-03-18T09:40:11.862,a,b2,c2,2020-03-18T09:40:11.862) 1> (true,a,b,c1,2020-03-18T09:40:11.862,a,b2,c2,2020-03-18T09:40:11.862) 1> (true,a,b,c,2020-03-18T09:40:11.862,a,b3,c3,2020-03-18T09:40:11.862) 1> (true,a,b,c2,2020-03-18T09:40:11.862,a,b3,c3,2020-03-18T09:40:11.862) 1> (true,a,b,c1,2020-03-18T09:40:11.862,a,b3,c3,2020-03-18T09:40:11.862) 1> (true,a,b,c,2020-03-18T09:40:11.863,a,b4,c4,2020-03-18T09:40:11.863) 1> (true,a,b,c2,2020-03-18T09:40:11.863,a,b4,c4,2020-03-18T09:40:11.863) 1> (true,a,b,c1,2020-03-18T09:40:11.863,a,b4,c4,2020-03-18T09:40:11.863) 1> (true,a,b,c,2020-03-18T09:40:11.863,a,b5,c5,2020-03-18T09:40:11.863) 1> (true,a,b,c2,2020-03-18T09:40:11.863,a,b5,c5,2020-03-18T09:40:11.863) 1> (true,a,b,c1,2020-03-18T09:40:11.863,a,b5,c5,2020-03-18T09:40:11.863) 1> (true,a,b,c,2020-03-18T09:40:11.864,a,b6,c6,2020-03-18T09:40:11.864) 1> (true,a,b,c2,2020-03-18T09:40:11.864,a,b6,c6,2020-03-18T09:40:11.864) 1> (true,a,b,c1,2020-03-18T09:40:11.864,a,b6,c6,2020-03-18T09:40:11.864){code} 第二次输出: {code:java} 1> (true,a,b,c,2020-03-18T09:42:36.168,a,b2,c2,2020-03-18T09:42:36.169) 1> (true,a,b,c,2020-03-18T09:42:36.171,a,b4,c4,2020-03-18T09:42:36.171) 1> (true,a,b,c,2020-03-18T09:42:36.171,a,b3,c3,2020-03-18T09:42:36.171) 1> (true,a,b,c1,2020-03-18T09:42:36.171,a,b2,c2,2020-03-18T09:42:36.171) 1> (true,a,b,c1,2020-03-18T09:42:36.171,a,b4,c4,2020-03-18T09:42:36.171) 1> (true,a,b,c1,2020-03-18T09:42:36.171,a,b3,c3,2020-03-18T09:42:36.171) 1> (true,a,b,c2,2020-03-18T09:42:36.172,a,b2,c2,2020-03-18T09:42:36.172) 1> (true,a,b,c2,2020-03-18T09:42:36.172,a,b4,c4,2020-03-18T09:42:36.172) 1> (true,a,b,c2,2020-03-18T09:42:36.172,a,b3,c3,2020-03-18T09:42:36.172) 1> (true,a,b,c,2020-03-18T09:42:36.188,a,b5,c5,2020-03-18T09:42:36.188) 1> (true,a,b,c2,2020-03-18T09:42:36.188,a,b5,c5,2020-03-18T09:42:36.188) 1> (true,a,b,c1,2020-03-18T09:42:36.188,a,b5,c5,2020-03-18T09:42:36.188) 1> (true,a,b,c,2020-03-18T09:42:36.188,a,b6,c6,2020-03-18T09:42:36.188) 1> (true,a,b,c2,2020-03-18T09:42:36.188,a,b6,c6,2020-03-18T09:42:36.188) 1> (true,a,b,c1,2020-03-18T09:42:36.188,a,b6,c6,2020-03-18T09:42:36.188) {code} -- This message was sent by Atlassian Jira (v8.3.4#803005) |
Free forum by Nabble | Edit this page |