Inserts data into a database table.
Two parameters may be provided to optimize
INSERT AS SELECT queries when
inserting out-of-order records into an ordered dataset:
batchCount(integer) value how many records to process at any one time
lagAmountwith a modifier to specify the unit of time for the value (i.e.
20sfor 20 seconds). The following table describes the units that may be passed:
unit description us microseconds s seconds m minutes h hours d days
Columns can be omitted during
INSERT in which case value will be
This method allows you to insert as many rows as your query returns at once.
Inserting out-of-order data into an ordered dataset may be optimized using
Using the lag and batch size parameters during
INSERT AS SELECTstatements is a convenient strategy to load and order large datasets from CSV in bulk. This strategy along with an example workflow is described in the importing data guide.
More details on ingesting out-of-order data with context on lag and uncommitted record count see the guide for configuring commit lag of out-of-order data
Hints and an example workflow using
INSERT AS SELECTfor bulk CSV import of out-of-order data can be found on the importing data via CSV documentation.