how works cache or persist with Spark Operator

[masalinas]

Please describe your question here

I'm using spark operator in minikube + minio to send some SQL distributed queries over CSV 2.4GB files with 8883 lines with 20000 columns each one and recovering 8883 samples with only two columns.

I would like cache or persist this queries. my question is, If the runners and driver are delete after fisnish, how can I cache or persist this queries from operator?