The following table lists the products and the products versions that you can directly upgrade to version 10.2.2 HotFix 1:

Product Version	Big Data Management	Enterprise Data Catalog	Enterprise Data Preparation
10.0	S	NS	NS
10.1	S	NS	NS
10.1.1	S	S	NS
10.1.1 Update 1	S	NS	NS
10.1.1 Update 2	S	S	NS
10.1.1 HotFix 1	S	S	NS
10.1.1 HotFix 2	S	NS	NS
10.2	S	S	S
10.2 HotFix 1	S	NS	NS
10.2 HotFix 21	S	NS	NS
10.2.1	S	S	S
10.2.1 Update 1	NS	S	S
10.2.1 SP1	S	S	S
10.2.1 SP2	S	S	S
S=Supported NS=Not Supported

Effective in version 10.2.2, mappings that use the Spark engine run in high precision mode by default.

You can disable the default setting at the mapping level and run the mapping in low precision mode by adding the following setting to mapping run-time properties:

Spark.SupportHighPrecision=FALSE

Set Spark.SupportHighPrecision to FALSE at the mapping level to run any mapping at low precision. You might want to do this to run mappings on the Spark engine that were developed in an earlier release and that you prefer to run with low precision. In some cases, such as a mapping in which scale is not specified, mappings fail at the default high precision setting and require low precision to run.