Github user tigerquoll commented on the issue:
https://github.com/apache/spark/pull/21306
So Kudu range partitions support arbitrary sized partition intervals, like
the example below, where the first and last range partition are six months in
size, but the middle partition is one year in size.
-- Make a table representing a date/time value as TIMESTAMP.
-- The strings representing the partition bounds are automatically
-- cast to TIMESTAMP values.
create table native_timestamp(id bigint, when_exactly timestamp, event
string, primary key (id, when_exactly))
range (when_exactly)
(
partition '2015-06-01' <= values < '2016-01-01',
partition '2016-01-01' <= values < '2017-01-01',
partition '2017-01-01' <= values < '2017-06-01'
)
stored as kudu;
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
