Class DbMerge<TInput>

Inserts, updates and (optionally) deletes data in database target. Before the Merge is executed, all data from the destination is read into memory. A delta table is generated that stores information if a records was inserted, updated, deleted or hasn't been touched (existed).

By default, identity columns (a.k.a auto increment or serial columns) are ignored when writing into the destination. If set to true, the DbMerge will try to keep existing identity values if possible.

The batch size used when inserted data into the database table.

The cache mode used for reading data from the destination table. By default, all data is read into memory before the actual merge is executed (Full mode). If set to partial, only data is read into memory that is needed for the current merge comparison. This will keep the memory consumption of the DbMerge much smaller.

Use a column converter to apply a conversion function to each value of a column before writing into the target

Column mapping to map property to column names. E.g. if the value of property "Id" should be written into the database column "Key", then you can simply add a column mapping: DbColumnName: "Id" -> PropertyName: "Key".

Property names that should be use to compare if the values of a column are equal, so that the DbMerge can decide if the column needs to be updated.

The connection manager used to connect to the database - use the right connection manager for your database type.

List of property names and a to-be value that tells the DbMerge if this row can be deleted.

The table definition of the destination table. By default, the table definition is read from the database. Provide a table definition if the definition of the target can't be read automatically or you want the DbMerge only to use the columns in the provided definition.

By default, the DbMerge expects the IdColumns to be unique. If another record with the same id values arrives, this record is then identified as a duplicate. It won't be written into the destination table - the delta output will contain this record, and the ChangeAction is set to Duplicate Please note that this may have an impact on the overall memory consumption, as already processed id values need to be stored internally during processing.

Property names that are used to check if the columns match (id values are equal).

If set to true, columns that have a default value declartions in the database are ignored when inserting data

Defines the type of target data which affects how deletions or insertions are handled. Full means that source contains all data, NoDeletions that source contains all data but no deletions are executed, Delta means that source has only delta information and deletions are deferred from a particular property and OnlyUpdates means that only updates are applied to the destination.

SourceBlock from the underlying TPL.Dataflow which is used as output buffer for the component.

The name of the target database table for the merge.

TargetBlock from the underlying TPL.Dataflow which is used as input buffer for the component.

Property names that describe which columns are actually updated (if an update of the row is necessary). Can be left empty, then all non id columns will be updated.

By default, the deletion of records are performed with a bulk delete. This performs well if the number of records to delete is small compared to the total amount of records. If this property is set to true, all records are deleted before the merge operation, using a TRUNCATE. Subsequently, all records are reinserted again. In order to make this work, the MergeMode needs to be set to Full and the CacheMode set to Full as well.

Exposes the ValueGeneratedColumns property from the underlying DbDestination used for inserts/updates.

If an error occurs in the component, by default the component will throw an exception and stop execution. If you use the error linking, any erroneous records will be caught and redirected.