Details
-
Test
-
Status: Resolved
-
Minor
-
Resolution: Incomplete
-
2.3.2
-
None
-
Important
Description
I have built Apache Spark v2.3.2 on Big Endian platform with AdoptJDK OpenJ9 1.8.0_202.
My build is successful. However while running the scala tests of "Spark Project REPL" module, I am facing failures at SingletonReplSuite with error log as attached.
The deviation observed on big endian is greater than the acceptable deviation 0.2.
How efficient is it to increase the deviation defined in SingletonReplSuite.scala
Can this be fixed?