Hi Youngwoo, What version of Flink and Json Format are you using? >From 1.11, we introduced `json.timestamp-format.standard` to declare the timestamp format. You can try `timestamp with local zone` data type with `ISO-8601` timestamp format.
Youngwoo Kim (김영우) <yw...@apache.org> 于2020年8月15日周六 下午12:12写道： > Hi, > > I'm trying to create a table using Flink SQL to query from a Kafka topic. > Messages from Kafka look like following: > > (snip) > "svc_mgmt_num":"7749b6a7e17127d43431e21b94f4eb0c116..." > "log_ymdt":"2020-08-15T02:01:33.968Z" > "snapshot_dt":"2020-08-13" > "network_type":"LTE" > > I'd like to make the column as 'TIMESTAMP' type but it does not work. Logs > from TM: > > Caused by: java.time.format.DateTimeParseException: Text > '2020-08-15T00:01:29.003Z' could not be parsed at index 10 > at java.time.format.DateTimeFormatter.parseResolved0(DateTimeFormatter > .java:1949) ~[?:1.8.0_265] > at java.time.format.DateTimeFormatter.parse(DateTimeFormatter.java: > 1777) ~[?:1.8.0_265] > ...... > > And also, If I declare the column with TIMEZONE (... log_ymdt TIMESTAMP > WITH TIME ZONE, ), the DDL does not work as well. > > [ERROR] Could not execute SQL statement. Reason: > org.apache.flink.sql.parser.impl.ParseException: Encountered "TIME" at > line 5, column 29. > Was expecting: > "LOCAL" ... > > How can I make my date-time format as the timestamp type? I'm running > Flink 1.11.1 and executing sql using FlinkSQL CLI. > > Thanks, > Youngwoo > -- Best, Benchao Li
