补充一下，flink版本是 1.17.1
在 2023-12-01 15:49:48，"casel.chen" <casel_c...@126.com> 写道： >线上有一个flink sql作业，创建和更新时间列使用的是 TIMESTAMP(3) >类型，没有配置'table.exec.sink.upsert-materialize'参数时是正常时间写入的`yyyy-MM-dd >HH:mm:ss.SSS`格式， >然后添加了'table.exec.sink.upsert-materialize'='NONE'参数后，输出的时间格式变成了 `yyyy-MM-dd >HH:mm:ss.SSSSSS`。数据类型变成了TIMESTAMP(6)，请问这是已知的issue么？ > > >-U[2023-11-29T21:11:02.327, 2023-11-29, 6666000017332097, 20231129, >YYHK650910000000000016607, S, 6666000017332097, 1006.50, 30, 04, 1, >23112921110248786000, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >+U[2023-11-30T12:43:04.676821, 2023-11-30, 6666000143554006, 20231130, >23113012430450887882, F, 6666000143718775, 10.00, 44, 07, 2, >23113012430450887895, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >+I[2023-11-29T17:37:01.556478, 2023-11-29, 6666000141180318, 20231129, >2f1edf1e3337642d, P, 6666000141538175, 246.00, 999, 01, 2, >23112917370147645164, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >-U[2023-11-25T16:02:45.145392, 2023-11-25, 6666000141288683, 20231125, >2023112516024553495256400285, P, , 1200.00, 81, 02, 1, 23112516024525664244, >2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >+U[2023-11-29T21:11:02.327, 2023-11-29, 6666000017332097, 20231129, >YYHK650910000000000016607, S, 6666000017332097, 1006.50, 30, 04, 1, >23112921110248786000, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >+U[2023-11-25T16:02:45.145392, 2023-11-25, 6666000141288683, 20231125, >2023112516024553495256400285, F, 6666000141586078, 1200.00, 81, 02, 1, >23112516024525664244, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >-U[2023-11-29T21:11:02.327, 2023-11-29, 6666000017332097, 20231129, >YYHK650910000000000016607, S, 6666000017332097, 1006.50, 30, 04, 1, >23112921110248786000, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683] >-U[2023-11-28T14:53:21.349043, 2023-11-28, 6666000137842973, 20231128, >HFPWALLET23112814532140921335, P, 6666000142774221, 62.98, 86, 06, 4, >538014532140921373, 2023-12-01T15:01:51.683, 2023-12-01T15:01:51.683]
