Doris Enhanced Features
Vectorized Computing
A vectorized engine is a query technology used by the data storage engine. Vectorization transforms operations that would traditionally be applied one-by-one to individual data elements (individualized operations) into a single operation applied to a "vector". Vector calculations are indeed implemented by leveraging specialized hardware within the CPU, specifically vector registers. This leads to faster query execution of databases. Thanks to the optimization of column-based storage and vectorized executor, the query execution speed will be improved by three to five times.
Column Pruning
Column pruning means that only required columns are read during query, and unnecessary columns are ignored, reducing data read overhead. When you query large tables containing numerous columns, you can use column pruning to exclude columns that are not required to reduce the amount of data to be read. This not only reduces the overhead of network transmission, but also reduces the overhead of storage on memory and disks. In addition, column pruning can improve the query response speed for better query performance, especially when processing massive data.
Pipeline Engine Optimization
The pipeline execution engine is built upon multi-core CPUs and is a data-driven execution engine. It fully leverages the computing power of multi-core CPUs, and limits the number of Doris query threads, resolving the problem of Doris execution thread bloat.
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot
