batch insert into aggregatingMergeTree table

[marlondu]

I have a table with aggregatingMergeTree engine, like:

CREATE TABLE db2.agg_dcc1_local ON CLUSTER ck_default_cluster( 
        id String,
        city String,
        codeUniq AggregateFunction(uniq,String), 
        codeCount AggregateFunction(count,String), 
        valueSum AggregateFunction(sum,Int32), 
        valueMin AggregateFunction(min,Int32), 
        valueMax AggregateFunction(max,Int32), 
        valueAvg AggregateFunction(avg,Int32), 
        create_time DateTime
)
ENGINE = AggregatingMergeTree() 
PARTITION BY toYYYYMM(create_time) 
ORDER BY (id,city)

So, How to batch insert data into the table with jdbc api ? I read ClickHousePreparedStatementImpl.java#executeBatch's code. because "insert into ... select ..." does not contains "VALUES", so it execute one by one. and it was too slow.

help. !!! my clickhosue version: 21.9.1.7426 clickhouse-jdbc: 0.3.2-patch8

[zhicwu]

Sorry @marlondu the driver does not support Aggregatefunction so you probably should use materialized view or projection etc instead.

As to batch insert, actually it's not that slow as it streams serialized data in RowBinary format.