Yes, you can perform incremental data load to merge the cache table with the most recent data retrieved from the source, I would perform the below steps:
* In Virtual DataPort Administration Tool, select your base view, click options and in Cache tab choose ‘Full’ cache mode.
* Check Incremental and add condition 'last_modified' > '@LASTCACHEREFRESH'. Ensure that the base view have a field 'last_modified' that indicates when a row was last inserted/updated. This condition will select rows from the original data source which have been added or updated since the last cache refresh. Save the base view.
* In Denodo Scheduler Administration tool , create a new VDPcache job to perform the cache load. To do so, you could refer to [Non Incremental Load Process](https://community.denodo.com/docs/html/browse/7.0/scheduler/administration/creating_and_scheduling_jobs/configuring_new_jobs/vdpcache_extraction_section#non-incremental-load-process).
* Choose data source, view and create a Load process as a Non-Incremental Load Process.
* For parameter, **Invalidate cache before load**, choose option ‘Matching rows’. This will invalidate rows that match the rows returned by the query before a cache load.
* Test and schedule this job based on your preferred time window under trigger section.
For more information, you could refer to the Knowledge Base article [Incremental Queries in Denodo](https://community.denodo.com/kb/view/document/Incremental%20Queries%20in%20Denodo?category=Performance+%26+Optimization).
Hope this helps!