How to size a Virtual Warehouse
When you create a Virtual Warehouse, you need to carefully set the size of your Virtual Warehouse. The size of the Virtual Warehouse you select during Virtual Warehouse creation determines the number of executors and concurrent queries the Virtual Warehouse can run.
Before creating a Virtual Warehouse and setting the Virtual
Warehouse size, learn about critical "Warehouse sizing requirements". If you
misconfigure the size of your Virtual Warehouse, you can use one of the following
methods, depending on the Impala or Hive type of Virtual Warehouse, to correct the
size:
When you create a Virtual Warehouse, you select one of the following Virtual Warehouse sizes:
| Virtual Warehouse Size | Number of Executors |
|---|---|
| XSMALL | 2 |
| SMALL | 10 |
| MEDIUM | 20 |
| LARGE | 40 |
| Custom | Enter a value between '1' and '100' |
If you are evaluating CDW, or just learning, XSMALL is the recommended size. For
production workloads, choose a size based on the following factors:
- The number of executors you typically use for clusters in an on-premises deployment.
- The complexity of your queries and the size of the data sets that they access.
Large warehouses with more executors can cache more data than small warehouses. Caching enhances performance.
