Deploying builds the selected tables in the Lake so they can load. Deploy from the Lake tree: mark a table, tick several items with Batch select, or right-click a datasource, database or schema to deploy the tables under it.
A table can be deployed when both of these are true:
Only tables are deployed. Selecting a datasource, database or schema collects the tables under it, and only the deployable ones are deployed; the rest are skipped and counted in the confirmation.
The dialog also warns that This action cannot be undone.
The button's tooltip changes with the selection:
| Tooltip | What it tells you |
|---|---|
| Select at least one deployable table | Select a table that can be deployed. |
| No tables in selection | The selection holds no tables. |
| All selected tables are already deployed | Every selected table is deployed already. |
| Selected tables are missing primary key | A selected table has no primary key. |
| No deployable tables in selection | Nothing in the selection can be deployed. |
| Deploy | The button is active; click it to open the confirmation. |
Right-click a row and choose Deploy. This is the menu item described in The Lake tree. It appears on a table that is not deployed and has a primary key, or on a datasource, database or schema that holds such tables.
The confirmation is titled Deploy {nodeType}: {name} and reads This will deploy the {nodeType} "{name}" to the Lake. Click Deploy to confirm.
Deploying builds the table; it does not load its rows. Loads run through the table's workflow, or you start one by hand. See Run, stop and troubleshoot loads.
When a deploy fails for some tables, a red badge appears next to Deploy with the number of failures. The count pulses for three seconds when a new failure arrives. Click the badge (its tooltip reads {n} table(s) failed - click to view) to open Failed Deployments:
#.Retries are per table, so you can fix and redeploy one table without touching the rest. For the message texts and the fixes behind them, see Lake troubleshooting.
Deploying needs Add/Modify Lake Objects. Without it, the request is refused. See Who can do what in the Lake.
It must not be deployed yet, and it must have at least one primary key column. Selecting a datasource, database or schema collects the tables under it, and only the deployable ones are deployed. Tables without a key are listed as skipped; to add one, see Configure a table.
The button activates only when the selection holds at least one deployable table. Hover over it to read the reason: No tables in selection, All selected tables are already deployed, Selected tables are missing primary key or No deployable tables in selection. Every message is listed in the table above.
The dialog counts the skipped tables by reason: {n} already deployed and {n} missing primary key. Already deployed tables need nothing. To deploy a skipped table that has no key, add a primary key on its Columns step and deploy it again. See Configure a table.
It counts the tables from your last deploy that failed. Click it to open Failed Deployments, where each table shows its error and offers Redeploy, Edit and Copy error. A table leaves the list once it deploys successfully.
Yes. Open Failed Deployments and click Redeploy on that table. Only that table is retried; a success removes it from the list, and a failure refreshes its error. You do not need to select or deploy the other tables again.
No. Deploying builds the table in the Lake; the load happens when the table's workflow runs, or when you start one by hand. See Run, stop and troubleshoot loads.
Add/Modify Lake Objects. Without it, the request is refused. See Who can do what in the Lake.
Yes. Right-click the datasource, database or schema and choose Deploy. The confirmation collects the deployable tables under it and lists the skipped ones. To hand-pick items instead, see Batch edit and organise.