SAP Signavio Targets for Replication Flows
If you use SAP Signavio as the target for your replication flow, you need to consider the following additional specifics and conditions.
General Information
-
Make sure you have created a Signavio connection in Connection Management and that it is available in your space. For more information, see SAP Signavio Connections
-
The target container is created automatically and called <datasphere space ID>/<Technical name of the replication flow>.
-
The default load type is Initial Only. You can change it to Initial and Delta (with a default time interval of 24 hours). Note that Delta Only is not supported.
Mask Sensitive Column Data
Additional Properties
The following properties are relevant for SAP Signavio targets. You can review these properties for the replication flow itself (by choosing ) or for individual replication objects (by selecting an object so that its properties get displayed in the side panel).
| Property | Value | Comment |
|---|---|---|
|
Group Delta by |
None |
It means that it is not possible to create additional folders for sorting updates based on the date or hour. |
|
File Type |
Parquet |
|
|
Enable Apache Spark Compatibility |
No |
It means that time data type columns are not converted to int64 and stored in the Parquet files. |
|
File Compression |
Snappy |
|
|
Suppress Duplicates |
No |
It means that if a data record already exists in the target, this record is still written to the target once again. |
Target Columns
|
Column |
Data Type |
|---|---|
|
__operation_type |
String |
|
__sequence_number |
Unit64 |
|
__timestamp |
Timestamp |
These columns are needed to capture changes in the source so that they can be replicated to the target, and you cannot change their names or settings.
If a target column has the same name as one of the columns for change data capturing, the target column has to be renamed. Auto-Projection does this by adding the prefix AUTOPREFIX_ to the name of the target column, for example AUTOPREFIX_timestamp.
Data Types
|
SAP Datasphere |
Parquet Type |
|---|---|
|
binary |
BYTE_ARRAY |
|
boolean |
BOOLEAN |
|
date |
INT32 |
|
decimal(p,s) |
FIXED_LEN_BYTE_ARRAY |
|
decfloat16 |
FIXED_LEN_BYTE_ARRAY |
|
decfloat34 |
FIXED_LEN_BYTE_ARRAY |
|
float32 |
FLOAT |
|
float64 |
DOUBLE |
|
int8 |
INT32 |
|
int16 |
INT32 |
|
int32 |
INT32 |
|
int64 |
INT64 |
|
string |
BYTE_ARRAY |
|
time |
INT64 |
|
timestamp |
INT64 |
|
uint8 |
INT32 |
|
uint64 |
INT64 |