Description

I am using JDBC driver 0.8.4 Spark3.5.1.

Write ClickHouse using Spark JDBC, Date type data, automatically converted time zones

My region is Singapore, spark job is configured as UTC, and Clickhouse time zone is also UTC.

However, after my data was converted to date type using the toDate function in Spark and written to Clickhouse,

The date type automatically treats Spark date data as Singapore time and then converts it again to UTC.

Actually, my Spark job is already UTC time.

JDBC does not require any additional parameter configuration.

Finally, I switched to JDBC version 0.2.6, and the data is now normal.

Environment

• Cloud
• Client version: 0.8.4
• Language version:Java11
• OS: Linux

ClickHouse Server

• ClickHouse Server version: 23.8