Hi, 

neither DLT stream, neither native Spark Structure Streaming, will not pick up a fact that record has changed. It can only read new comings data. 

1. If you want to maintain incremental loading of data, and you want to read data which are added, remove this option from you pipeline 

.option("readchangeFeed","true")

and check if your pipeline works fine, by adding additional file to this location: 

path="/mnt/saphana-adls-landing/saphana-adls-landing/customer_landing"

2. If you don't care about incremental loading of data, but you care about data being changed, you can do the full reload, by changing 

spark.readStream.format("cloudFiles")
      .option("cloudFiles.format", "csv")

to:

spark.read.csv()

3. There is also something called Change Data Feed, but its more advance, and I dont think that its what you are looking for. You are read more about it here: https://docs.databricks.com/en/delta/delta-change-data-feed.html

Good Luck