FAQs
Specific
Is incremental indexing supported by the connector?
Yes. Currently, this is done by accessing SharePoint Online changes database directly, so after a full scan the connector will only fetch objects that change in a given date range (since the last scan and the current scan).
SharePoint Online doesn't keep track of item updates/deletes for BCS external list items on its change log. To be able to check for changes on BCS external lists, all external list items are crawled and checked against the information on the current snapshot file. If any of those items has any changes, then it is sent to the pipeline to be updated or deleted.
For a discussion on crawling, see here.
Why does hierarchy differ when using snapshots in crawls?
Snapshots store the full status of items during full crawls. By contrast, not using snapshots means that change information will be taken from the SharePoint change log, which in some cases doesn't provide all information such as deletes. This behavior is expected.