You can override default Hyperscale settings for individual tables within a rule set to fine-tune performance.
When to use this task
Use this task when specific tables require different Hyperscale processing parameters than the job-level defaults. For example, large tables that need more parallel connections or different chunk sizes.
Steps
-
Identify the rule set containing the tables you want to configure.
-
Update the table metadata for the rule set, specifying the Hyperscale attributes to override for each table. In the DCT UI, table-level overrides can also be configured during the Enable and Run Hyperscale Job wizard on the Target Configuration step, where each table is listed with its current values and an edit icon to modify individual settings. A Reset All To Default button is available to revert all table-level overrides to the orchestrator defaults.
Configurable table-level attributes include:
| Attribute | Description |
|---|---|
unload_split
|
Number of splits used when unloading data from the source table. |
stream_size
|
Size of the data stream buffer for the table. |
max_parallel_connections_per_table
|
Maximum number of parallel connections allowed per table. | |
max_records_per_split
|
Maximum number of records processed per split. |
max_split_per_connection
|
Maximum number of splits handled by a single connection. |
column_array_rows
|
Number of rows fetched per array operation. |
auto_calculate_unload_split
|
When enabled, automatically calculates the optimal unload split value. |
filter_key
|
Column used to partition data during unload |
What's next