How to increase pentaho transformation input/output speed

I am using Pentaho data integration (PDI)-spoon to create ETL's and I am very focused on performance.
I develop an ETL to process that copy of 2,500,000 rows (each row has 104 columns) from MySQL 8 to Clickhouse database and it takes 30 min.
Destination table does not have any indexes and constraints and it is a columnar database.

I am using linux ubuntu 22.04 and transformation running on pentaho server through spoon.sh

How to increase the transformation input/output speed?

I am using only 4 steps:-
Truncate table by using EXECUTE SQL SCRIPT --> Fetch data by using TABLE INPUT--> Changing date formats by using SELECT VALUES ---> insert data into destination table by using TABLE OUTPUT.

I want to increase the I/O speed of the PDI-Spoon transformation

If it's a point-in-time migration, uou can try using the MySQL Table engine in ClickHouse in conjunction with INSERT FROM SELECT syntax to migrate your data directly from your ClickHouse instance, it should be faster.

https://clickhouse.cloud/integrations/mysql
https://clickhouse.com/docs/en/engines/table-engines/integrations/mysql

There are couple of options in PDI to improve the I/O:

1. Use of Carte Cluster.

You can make use of multiple carte servers and execute the job to run in parallel across the cluster. The way to create carte cluster and execute them is written in this blog.

2. Using Bulk Loader step

You can use the Bulk Load from MYSQL to File step in PDI to bulk load the data into a file. This step will improve your I/O. You can later use native clickhouse database APIs to load the data from file to table.

Alternatively, you can also use Text File Output to write the data into a file and upload the data into the clickhouse database.