Staging
Staging is a feature for our Tabular Storage controller.
Tip
When ingesting large datasets, it makes sense to do it in chunks - instead of one go.
Staging data with the SDK
When a tabular dataset has been created, and a schema has been made, data can be uploaded using staging.
First, we need to create a stage request with the create_stage_request
function.
We are then returned an ID for our stage.
To upload data to the stage we use the write
function - which optionally accepts a table_stage
parameter, this is where we'll put the stage id we received.
When we are done writing data to the stages we have created, we can commit_stage_request
with table_stage
for the stage.
This tells the platform that you are finished uploading data to the stage, and you want to load everything in the stage over to the dataset.
Follow the progress by polling list_stage_request
for the dataset, or get_stage_request
for each stage by passing the table_stage_identifier
.
This way we can ask how the stages are going. When their statuses are all "completed" they are ingested to the dataset.