Manage connections
Once you have created or edited one or more connections, you can manage them in Connections. Connections let you:
- View all your connections at a glance, including the owner, the sandbox, and when the connections were created and modified.
- Edit a connection.
- Delete a connection.
- Create a data view from a connection.
- View all datasets in a connection.
- Check the status of your connection’s datasets and the status of the ingestion process. For example, when is your data available so that you can start with reporting and analysis in Analysis Workspace.
- Identify any data discrepancies due to misconfiguration. Are you missing any rows? If so, what rows are missing and why? Did you misconfigure connections and cause missing data in Customer Journey Analytics?
A table shows all connections available. You can quickly search for a connection using the Search
The following columns / icons are available in the table.
To view information about Datasets included, Sandbox, Owner, and more, select
A popup window displays details.
Select
Shows one or more links to the datasets that are part of the connection. You can select the dataset hyperlink to view the dataset in the connection. If more datasets are part of the selected connection, select +x more to show a Datasets included panel. This panel shows links to all datasets and an option to search for a specific dataset that is part of the connection.
Selecting a dataset name opens the dataset in the Experience Platform UI in a new tab.
Shows the status of importing new data for datasets:
● x On for how many datasets are configured to import new data, and</
● x Off for how many datasets new data import is turned off.
Shows the status for backfill data across datasets.
● x backfills failed for number of failed backfills across datasets,
● x backfills processing for number of processing backfills across datasets,
● x backfills completed for number of completed backfills for datasets, and
● Off in case no backfills are defined for the datasets in the connection.
You can configure which columns to display using
Edit a connection
Allows admins to edit the connection.
- Select
- Select
Alternatively, you can:
-
Select the connection row.
-
Select
When editing a connection, you can:
- Start and stop importing new data.
- Rename a connection.
- Refresh the dataset/s.
- Remove dataset/s from the connections.
See Create or edit a connection for more information.
Delete a connection connections-delete
Allows admins to delete the connection.
- Select
- Select
Alternatively, you can:
-
Select the connection row.
-
Select
When you delete a connection, a Delete connection panel indicates which data views are deleted and which workspace projects are affected.
Select Continue to delete the connection.
See Delete implications for more information about the implications of deleting a connection.
Create a data view
Allows admins to create a data view for the connection.
-
If no data view is associated with the connection:
- Select
- Select
-
If one or more data views are already created for the connection:
- Select
- Select
- Select
Alternatively, you can:
-
Select the connection row.
-
Select
See Create or edit a data view for more information.
Connection details connection-detail
To go to the details for a connection, select a connection name in the connections table.
The Connections details interface provides a detailed view of the status of a connection. You can:
- Check the status of your connection’s datasets and of the ingestion process.
- Identify configuration problems that can cause skipped or deleted records.
- See when the data is available for reporting.
Summarizes the event records added/skipped/deleted, and the number of batches added, for the dataset and date range you have selected.
Select Check detail to show the Check skipped detail popup, listing for all event datasets or selected dataset the number of skipped records and the reason.
Select
Indicates how many rows were skipped in the selected time period, for the dataset and date range you have selected. Reasons for skipping records include: missing timestamps, missing or invalid person ID, and so forth. Updated every 10 minutes.
Invalid person IDs (such as “undefined”, or “00000000”, or any combination of numbers and letters in a Person ID that appears in an event more than 1 million times in a given month) cannot be attributed to any specific user or person. They cannot be ingested into the system and result in error-prone ingestion and reporting. To fix invalid person IDs, you have 3 options:
- Use Stitching to populate the undefined or all-zero user IDs with valid user IDs.
- Blank out the user ID, which will then be skipped during ingestion (preferable to invalid or all-zero user IDs).
- Fix any invalid user IDs in your system before ingesting the data.
Shows the status of importing new data for the dataset:
● x On if dataset is configured to import new data, and
● x Off if dataset is configured to not import new data import.
Shows the status of backfill data for the dataset.
● x backfills failed for number of failed backfills,
● x backfills processing for number of backfills processing,
● x backfills completed for number of backfills completed, and
● Off in case no backfills are configured.
Connection panel
When no dataset is selected in the datasets table, a panel on the right side of the Connections interface shows connection options and details.
Shows the status of importing new data for datasets:
● x On for how many datasets are configured to import new data, and
● x Off for how many datasets new data import is turned off.
Shows the status of backfill data for datasets.
● x backfills failed for number of failed backfills across datasets,
● x backfills processing for number of processing backfills across datasets,
● x backfills completed for number of completed backfills for datasets, and
● Off in case no backfills are defined for the datasets in the connection.
Dataset panel
When a dataset is selected in the datasets table, a panel on the right side of the Connections interface show details for the selected dataset.
Indicates how many rows were skipped during ingestion in the selected time period.
Reasons for skipping records include: Missing timestamps, missing or invalid person ID, and so forth. Updated every 10 minutes.
Invalid person IDs (such as “undefined”, or “00000000”, or any combination of numbers and letters in a Person ID that appears in an event more than 1 million times in a given month) cannot be attributed to any specific user or person. They cannot be ingested into the system and result in error-prone ingestion and reporting. To fix invalid person IDs, you have 3 options:
- Use Stitching to populate the undefined or all-zero user IDs with valid user IDs.
- Blank out the user ID, which is then skipped during ingestion (preferable to invalid or all-zero user IDs).
- Fix any invalid user IDs in your system before ingesting the data.
Shows the status of importing new data for the dataset:
● x On if dataset is configured to import new data, and
● x Off if dataset is configured to not import new data import.
Shows the status of backfill data for the dataset.
● x backfills failed for number of failed backfills,
● x backfills processing for number of backfills processing,
● x backfills completed for number of backfills completed, and
● Off in case no backfills are configured.
To show a dialog with an overview of the past backfills for the dataset, select
