大佬们麻烦帮忙看下这个 CK 查询速度为啥比 MySQL 慢

单表两千万数据，MySQL 4C8G,24ms ，clickhouse 2C4G,282ms

MySQL:

SELECT
corp_id,
corp_user_id,
SUM(totalxx_num) totalxx_num,
SUM(total_xxx_num) total_xxx_num
from xxx.xxxx
where corp_id = 'sdd' and dept_id = 100 and send_day >= '2023-03-01 00:00:00'and send_day <= '2023-04-25 00:00:00'
group by corp_id,corp_user_id;

CK:

SELECT
corp_id,
corp_user_id,
SUM(totalxx_num) totalxx_num,
SUM(total_xxx_num) total_xxx_num
from xxx.xxxx
where corp_id = 'sdd' and dept_id = 100 and send_day >= toDate('2023-03-01 00:00:00') and send_day <= toDate('2023-04-25 00:00:00')
group by corp_id,corp_user_id;

DDL:

create table xxxx
(
    id                          Int64,
    corp_id                     String,
    user_id                     Int64,
    corp_user_id                String,
    totalxx_num              	Int32,
    total_xxx_num               Int32,
    send_day                    Date
)
    engine = MergeTree PARTITION BY toMonday(send_day)
        PRIMARY KEY id
        ORDER BY (id, corp_id, user_id, send_day)
        SETTINGS index_granularity = 8192;