Batch edit changes the settings of many Lake tables from one datasource in a single grid. Use it when many tables need the same sync type, permission groups or incremental settings. Batch select lets you deploy or delete several tree items at once, and folders keep a long datasource list tidy.
Batch Edit is available for a datasource that is ready. Open it in any of these ways:
It opens the Batch Edit Tables grid with every table of the datasource. If the datasource is not ready, the toolbar button is greyed out and its tooltip reads Select a ready datasource to batch edit, and the right-click menu does not list Batch Edit.
The grid shows Loading tables while it loads. If the load fails, it shows Failed to load tables; close the popup and open it again. A datasource without tables shows No tables found with This data source has no tables to configure., and the grid itself reads No tables found.; import tables from the source first, see Import tables. If the popup cannot tell which datasource to use, it shows Error: No data source selected with a Close button.
Your unsaved edits survive a page refresh and stay available for a day, so you can come back to a batch you started earlier.
The grid lists every table of the datasource with its catalog details, sync settings, incremental settings, criteria and query settings. The Batch edit grid describes every column and cell.
A change copies to the other ticked rows only when the row you edit is ticked and at least two rows are ticked. A ticked row that cannot take the change keeps its value. Incremental Column, Minus Day and Full Reload copy only to rows with an incremental sync type, and Loop Increment only to rows with the loop sync type. Apply to filtered rows, under Permission Groups, sets the groups on every row that matches the current filters.
Changing Sync Type also clears the fields the new type does not use. When the new type needs another kind of workflow, for example a Snapshot workflow, the table moves to the first workflow of that kind on the datasource, or is left without a workflow if there is none. Database, Schema, Table Name and Row Count are read-only. A locked cell says why, for example Available only for incremental sync types. The sync types are described on Choose a sync type.
When you click Save, the grid checks every changed row first; the tables you are removing are not checked. If any changed row fails, nothing is saved. A Validation Error message lists the problems, each prefixed with the table name:
| Message | What to do |
|---|---|
| Workflow name is required when synchronize is enabled | The table has no workflow. Turn Synchronize off for it, or add a workflow of the matching type and set it as the table's Workflow Name on Table > General. See Workflows in the Lake. |
| Incremental column is required | Choose an Incremental Column. |
| Minus day must be set | Enter a Minus Day of 0 or more. |
| Loop increment must be at least 1 | Enter a Loop Increment of 1 or more. |
| Full reload start date is required for the Loop sync type when Full Reload is enabled | Set Full Reload Start Date, or turn Full Reload off. |
Once the check passes, the result appears as a notification:
| Notification | Meaning |
|---|---|
| Saved | All changes were saved. The message reads "All N table(s) updated successfully." |
| Partially saved | Some tables were saved. The message reads "X of Y table(s) updated. Check the error bar for details." Failed rows stay in the grid with their changes; the rows that saved reset. |
| Save failed | No table was saved. The message reads "No tables were updated. Check the error bar for details." Your changes stay in the grid. |
| No changes | There was nothing to save. The message reads "There are no changes to save." |
When tables fail, a button showing the number of failed tables appears in the toolbar. Click it to open Failed Updates. It lists each table with its error, and its hint pulses for three seconds when new failures arrive. Retry re-runs one table's save, Edit opens the table and closes the list so you can fix it and save again, and Copy error copies the message and briefly shows a check. Tables that saved are no longer in the list.
One save carries at most 500 updated tables and 500 removals. If a table is both changed and removed in the same save, it is removed.
Click Revert to discard all unsaved changes. Save and Revert stay greyed out until you change something.
Removing a table deletes it from the Lake together with its loaded data. Save does not ask you to confirm. Click Revert before you save if a row should stay.
A table that other tables depend on, or that a Writeback table uses as its source, cannot be removed. It stays in the grid and is listed under Failed Updates. A table that a Connect Any API connector reads through a data-access grant is blocked as well; the reason on the table names the connector and the grant to revoke before you try again. See What uses a table.
The Status column shows one of three badges. A Deployable table opens Deploy table? when you click it, and its tooltip reads Click to deploy this table; confirm with Deploy to build the table in the Lake with its current column configuration. A Missing PK table opens it on the Columns step, so you can define a primary key. A Deployed table is locked here and cannot be redeployed from this grid.
To deploy from the tree instead, see Deploy tables.
Leaving batch select clears the selection. While it is on, View Source is disabled and its tooltip reads Disabled in batch select mode.
Deploy opens Deploy Selected Items?, which shows Tables to deploy and any skipped tables: those already deployed and those with no primary key. The action cannot be undone. Delete opens Delete Selected Items?, which first checks what depends on the selection; if something does, the confirm button reads Delete anyway. For the full delete flow, see Delete Lake items.
Failed deployments and deletions appear behind a button showing the number of failures. Each failed table offers Redeploy or Retry delete.
A folder name must not be empty or match another folder's name, letter case aside; until the name is usable, Create stays off and the panel answers Name is required or A folder with this name already exists. If the request fails, the message reads Failed to create folder.
To move a datasource into a folder, drag it onto the folder, or open its connection form and choose a Folder. Dropping a datasource back onto the folder it is already in changes nothing; drop it on empty top-level space to take it out of a folder. A move is confirmed with DataSource moved successfully; a failed one reads Failed to move datasource.
A folder can be deleted only when it is empty. Its menu offers only Delete, and while it holds datasources the tree refuses with Cannot delete folder and Please remove or move all data sources from this folder first. An empty folder is removed with Folder deleted successfully. A folder keeps its colour and appears in the tree's Folder filter once it holds a datasource; see Search and filter the Lake.
Saving a batch edit, deploying tables and moving a datasource between folders need Add/Modify Lake Objects. Without it, the server refuses the save. See Who can do what in the Lake.
Yes. Open Batch edit, tick the 40 rows and change Sync Type on one of them to Incremental Reload (Truncate Insert). The change copies to the other ticked rows. Then set Incremental Column and Minus Day on the ticked rows and click Save. A row that does not have the column you chose keeps its value, so check the rows before you save.
Minus Day applies only to rows with an incremental sync type. The other rows lock that cell and keep their value.
One changed row failed validation, so no row was saved. Read the Validation Error message, which names each table, and fix every listed row. Then click Save again.
Click the button that shows the number of failed tables in the toolbar, then open Failed Updates. Retry re-runs a table's save, or click Edit to fix the table and save again.
Click the table's Missing PK badge in the Status column. The table opens on the Columns step, where you define a primary key and click Apply. See Configure a table.
Yes. Tick the tables, click Remove selected tables, then click Save. Removal deletes each table and its loaded data, and Save does not ask you to confirm. Tables that other tables depend on, or that a Writeback table uses as its source, stay in the grid under Failed Updates.
Yes, with Batch select in the tree. Tick the items and click Deploy. Tables that are already deployed or have no primary key are skipped, and the dialog shows how many. You can also deploy one table at a time from the grid's Status column.
Yes. Unsaved edits stay as you left them when you refresh the page, and they remain available for a day.
The datasource has no tables yet, so the grid shows No tables found with This data source has no tables to configure. Import tables from the source first. See Import tables.
Not while it has datasources. The folder menu offers only Delete, and a folder with datasources cannot be deleted. Move the datasources out first. To change a name, create a new folder, move the datasources into it and delete the empty one.
Batch edit works only on a datasource that is ready. Select a datasource that shows as ready. If the datasource is stuck in another state, see Lake troubleshooting.