Back
Step 3: Running a Connector
Troubleshoot connector result previews, slow or stuck connectors, and specifics of different run modes.
Preview Table
Why are all my rows filtered out?
The preview table only displays a small subset of the data to be ingested by the Connector. The preview table supports pagination, but the next page is not pre-fetched until explicitly asked to do so.
This means that all transforms, specifically the Filter Rows transform, are applied to only the rows on the given page of results.
For example, if the Filter Rows transform happens to filter out all the rows on the first page, but not the rows on subsequent pages, you will need to explicitly paginate to the next page to view the next set of rows.
Run Progress
Why is my Connector stuck?
Most likely, the connector is moving slowly, but the run is still in progress. Delays in connector processing could be due to a few different reasons:
- Increased traffic on the Connectors system as a whole
- Large volumes of entity data
- Using a sandbox account vs. a production account (connector processing takes more time in sandbox environments)
The progress bar reports on each step as follows:
| Progress Percentage | Stage | Description |
|---|---|---|
| 0-70% | Pre-processing | All data extraction and transformation steps |
| 71-98% | Loading entities | The entity job to load any new entities or entity updates into the Knowledge Graph |
| 99%-100% | Completing the run | Marking the connector run as “complete” and generating the downloadable results file. This stage takes significantly longer in sandbox environments than in production. |
It’s important to note that certain stages of the run process can take longer than others (e.g., if a connector’s progress has been at 70% for a long time, this may mean that it is taking longer to load the data updates than it did to extract the data).
Why is Run Results data not loading?
Depending on the amount of traffic at a given time, results may take some time to load.
Run Modes
For more on run modes, see these references:
Why were some entities not deleted if they were missing from a run on Comprehensive Mode?
Unless the comprehensive data set is defined with a saved filter (and the missing entities are present in that filter), only entities that were originally created by the connector will be considered part of the comprehensive set and deleted if missing from a run on comprehensive mode.
Why are the Dry Run results different from the actual Run Results on the same connector?
This can happen if any changes to entity data were made between when the dry run was completed and when it was approved.
When a dry run is approved, entity changes are executed based on the ingested data at the time of the dry run (so the approval does not re-fetch data or re-run the connector).
For example, if Entity A would have been updated according to the dry run, but was then deleted from the Knowledge Graph, and the dry run was then approved, Entity A would not be present in the actual run results of the connector.
Why is the Yext ID field blank after a Dry Run?
Yext IDs are only assigned at the time of entity creation, and dry runs do not create entities.