Hi,
The message informs you that your project contains more than 100,000 results lines, and that you will only be served results files up to 100,000 lines. The solution is to download the results in multiple batches, only including series of submissions that total fewer than 100,000 results lines at a time (refresh the page after you download a batch, as sometimes memory gets clogged). Then you can concatenate the multiple CSV files that you’ll have downloaded into a single one on your local machine if you like (on Unix systems, you can do it very quickly with a command like cat results*.csv > results_all.csv)
I recommend you unpublish your experiment when you do that, to avoid missing new incoming data in the process, and once all your results have securely been downloaded to a safe location, delete them from the farm to free up space in our database, to make saving and accessing future data faster, and to avoid facing this situation again. (Remember to publish the experiment again if you plan on collecting more data)
Jeremy