We are starting a project with PostgreSQL and will need to use Citus in the near future for a multi tenant application so we are aiming at preparing our PostgreSQL database schema appropriately for an easy upgrading to Citus. I was reading the following page https://learn.microsoft.com/en-us/azure/postgresql/hyperscale/concepts-choose-distribution-column#best-practices and it states the following:
"Partition distributed tables by a common tenant_id column. For instance, in a SaaS application where tenants are companies, the tenant_id is likely to be the company_id."
The question is whether the term "Partition" in the statement from above is referring to PostgreSQL table partitioning (https://www.postgresql.org/docs/14/ddl-partitioning.html) or is it referring to Citus sharding by key? Does PostgreSQL table partitioning by tenant_id make any sense or provide any benefit when sharding table in Citus with the same sharding key as the one used to partition table in PostgreSQL (tenant_id)?