Teknik Subpartisi

Welcome! In this lesson, we'll explore sub-partitioning, an advanced technique to combine different partitioning methods in PostgreSQL.

It allows you to organize your data with even finer granularity, creating a powerful hierarchical structure for very large tables.

Sub-partitioning offers several key advantages for managing and querying massive datasets:

• Finer Granularity: Break down large partitions into smaller, more manageable units.
• Targeted Management: Easier to perform operations (e.g., attach, detach, archive) on specific data subsets.
• Improved Query Performance: The database can prune even more irrelevant data blocks, significantly speeding up queries on specific sub-sections.

With sub-partitioning, you define a primary partitioning strategy for your main table. Then, for each individual partition of that main table, you define a secondary partitioning strategy.

Think of it as partitioning a table by year, and then partitioning each year's data further by region. It's a 'partition of a partition' concept.

A common and effective sub-partitioning pattern is to first partition a table by a date range (e.g., year or quarter), and then sub-partition each date range by a list of discrete values (e.g., region, department, status).

This is ideal for time-series data that also has important categorical attributes, allowing you to quickly filter by both.

Let's create an orders table. This will be our top-level parent, partitioned by order_date using RANGE partitioning.

Now, we create a partition for the year 2023. Crucially, we add PARTITION BY LIST (region) to this partition definition.

This makes orders_2023 itself a parent table, ready for its own sub-partitions.

Finally, we create the actual sub-partitions for specific regions within the orders_2023 partition. Data for 'North' goes into orders_2023_north, etc. We'll also insert some data to see it in action.

You can also reverse the strategy: partition first by a list of categories (e.g., 'Electronics', 'Books'), and then sub-partition each category by a date range (e.g., release year).

This is useful when your primary access pattern is by category, and then you need to filter within categories by time.

Sub-partitioning offers powerful ways to organize data. Which of the following statements correctly describe its characteristics or benefits?

You've now learned about PostgreSQL sub-partitioning!

• We saw how to combine RANGE and LIST partitioning to create deeply organized tables.
• This technique provides finer data granularity and can significantly boost query performance by enabling more precise partition pruning.
• Understanding sub-partitioning is crucial for managing extremely large and complex datasets effectively.

Next, we'll dive into managing partitioned tables, including adding, dropping, and altering partitions efficiently.