Creating migration tasks
Migration tasks contain the lists of tables to be migrated.
The Data Migration module provides for creating migration tasks for:
- Schema
- Tables node
- Separate table
- Several tables
- Several tables from different schemas
Creating a migration task
- Check the schema in the Source metadata tree
- Click the Create migration task button. (This option is also available in the module toolbar and in the source tree context menu.)
Figure 1. Create new task buttons
The matching process begins and depending on the matching results the Task Options or the Data Objects tab opens.
If some data objects left unmatched the Data Objects tab opens (Figure 2). In this case, you have to:
- remove the unmatched objects from the task, or
- add the matching pair manually
Note: The task can be saved when data objects contain valid pairs only.
Figure 2. Data objects tab
The data Objects tab divides pairs into the following groups:
- Valid — matched tables from the Source and Target sides
- Pending — matched tables from the Source and Target sides that may differ in a datatype, quantity of columns, etc.
- Not matched — appropriately matched tables not found on the Target side
If the table has partitions, the number of partitions is shown at the end of the row.
If the matching process went successfully and the data objects contain only valid pairs the Task Option tab opens. In the Task Option tab, you set the number of data streams, set the truncate options, and specify the number of attempts to connect to the server in case if it’s unavailable.
When all the unmatched pairs are resolved (and nothing highlighted red appears in the Data objects tab) you can save your migration task.
- Click the Save button. When created a task appears in the Migration Tasks field and the Processing info tab opens.
- Click the Start button in the Processing info tab (Figure 3) to start the migration task.
Figure 3. “Processing info” tab
Column mapping options
You can configure column mapping for each pair. For that use button in form gear wheel of pair (available for valid and pending pairs only). Check Figure 4 to see an example of the Column mapping options form.
Figure 4. Column mapping options form
In this form, you can:
- Exclude columns from migration by set/unset checkboxes;
- Change a column destination on the Target side;
- Set a custom transformation for the Source column before migration to the Target side.
Note: Exclude columns and Custom transformation options are available in the Set migration form that performs the same logic. When settings in the Column mapping options form have a higher priority.
You can remove the whole list of invalid objects. Invoke context menu in the form and chose to Remove all invalid items (Figure 5).
Figure 5. Removing all invalid objects
Task options
The Task options tab (Figure 6) allows you to select a template for the task and set the task parameters.
Figure 6. Task options tab
The Task options tab contains the following fields:
- Name — a task name, can be non-unique, non-empty (100 symbols);
- Description — a short task description (255 symbols);
- Templates — a migration template (check the supported templates in the context of dialects in the Templates page);
- Source DB — an information block that contains connection parameters of the Source DB, not for editing;
- Migrator — the task settings:
- Recreate constraints: used for recreating constraints in the Target tables. If the checkbox is set, all the constraints should be deleted/disabled in the Target tables before migration. Be aware that this action might affect tables that are not taking part in migration on the Target side;
- Disable triggers: used for disabling triggers in the Target table. If the checkbox is set, all the triggers should be disabled in the Target table before migration. After the migration is performed, triggers are tried to enable. In some cases, it is impossible to disable/enable triggers, then you need to analyze error logs. Note: This setting isn’t displayed for Oracle-MSSQL pairs due to the BULK operation which already includes the disable/enable triggers;
- Truncate tables: used for truncating the Target table before migration;
- Reader batch size: used for defining the size of the reader batch in raws, the default value is 100000;
- Writer batch size: used for defining the size of the writer batch in raws, the default value is 10000;
- Writer timeout: used for defining the timeout of the writer operation in seconds, the default value is 30000;
- Attempt count: a number of attempts for the migration when the operation failed, the default value is 3;
- Enable staging table: used when migration is performed with staging tables. In this case, the migration is carried out in the staging (temporary) tables, and not directly to the Target table. After migration, all the data is copied to the Target table, the staging tables are deleted. It allows for avoiding errors during the migration of big tables. Note: This setting works for Oracle-MSSQL pairs, and when the table has partitions or filters.
- Staging tables schema: used for defining the schema name for the staging tables. If the field is empty, staging tables will be created in the schemas of the corresponding Target tables;
- Target DB: an information block that contains connection parameters of the Target database, not for editing.
After specifying the options of the task and objects for migration, click the Save button.
The Task options tab has additional actions after saving:
- Save mapping — save data objects pairs of a task in a JSON file;
- Save template — save settings of a task to a JSON file.
To modify the migration settings for a particular table select the Set migration option in the source tree context menu. See the details on the Migration options page.
Task actions
The task actions are presented on the Task actions page.
Task statuses
The task statuses are presented on the Task statuses page.
Didn’t find the answer?
You can report problems, ask questions or share ideas for improvements on our email [email protected].