What Are the Differences Between MRS Spark and DLI Spark?

Both DLI and MRS support Spark, but there are some differences in service mode, interface, application scenarios, and performance characteristics.

• The Spark component of DLI is fully managed. You can only use it through its APIs.

  This mode reduces the O&M burden and allows you to focus more on data processing and analysis tasks.

  For details, see Data Lake Insight Developer Guide.

• The Spark component of MRS is built on the VM in an MRS cluster. You can adjust and optimize Spark according to your actual needs and make API calls to use it.

  This mode provides higher freedom and customization, and is suitable for users with experience in big data processing.

  For details, see MapReduce Service Developer Guide.