Migrating from Redshift to ClickHouse

Both databases are column-based, they excel at handling a large number of columns and are very good at compressing data on disks (in Redshift, you can even configure the compression types for each specific column). One day, one table

Data are sorted on the disk and deleted in Redshift when you run the following command: VACUUM

The vacuum process works with all data in the table. After spending a couple weeks testing a small data copy, we understood that some issues needed to be solved before fully replacing Redshift with Clickhouse:

What types of instances and disks should we use for deployment?

Source: www.altinity.com