📄️ Gather statistics for CBO

This topic describes the basic concept of StarRocks cost-based optimizer (CBO) and how to collect statistics for the CBO to select an optimal query plan. StarRocks 2.4 introduces histograms to gather accurate data distribution statistics.

📄️ Synchronous materialized view

This topic describes how to create, use, and manage a synchronous materialized view (Rollup).

🗃️ Asynchronous materialized views

4 items

📄️ Colocate Join

For shuffle join and broadcast join, if the join condition is met, the data rows of the two joining tables are merged into a single node to complete the join. Neither of these two join methods can avoid latency or overhead caused by data network transmission between nodes.

📄️ Use Lateral Join for column-to-row conversion

Column-to-row conversion is a common operation in ETL processing. Lateral is a special Join keyword that can associate a row with an internal subquery or table function. By using Lateral in conjunction with unnest(), you can expand one row into multiple rows. For more information, see unnest.

📄️ Query Cache

The query cache is a powerful feature of StarRocks that can greatly enhance the performance of aggregate queries. By storing the intermediate results of local aggregations in memory, the query cache can avoid unnecessary disk access and computation for new queries that are identical or similar to previous ones. With its query cache, StarRocks can deliver fast and accurate results for aggregate queries, saving time and resources and enabling better scalability. The query cache is especially useful for high-concurrency scenarios where many users run similar queries on large and complex data sets.

📄️ Data Cache

From v3.1.7 and v3.2.3 onwards, StarRocks introduced Data Cache to accelerate queries in shared-data clusters, replacing File Cache in earlier versions. Data Cache loads data from remote storage in blocks (on the order of MBs) as needed, while File Cache loads entire data files each time in the background, regardless of how many data rows are actually needed.

🗃️ Computing the number of distinct values

2 items

📄️ Sorted streaming aggregate

Common aggregation methods in database systems include hash aggregate and sort aggregate.

📄️ Accelerate COUNT(DISTINCT) and Joins with AUTO INCREMENT and Global Dictionary

This topic describes how to accelerate COUNT(DISTINCT) calculation and Joins using AUTO INCREMENT columns and Global Dictionary.

📄️ [Preview] Flat JSON

This topic introduces the basic concepts of the Flat JSON feature and how to use this feature.