SPARK-2757 [BUILD] [STREAMING] Add Mima test for Spark Sink after 1.10 is released

Re-enable MiMa for Streaming Flume Sink module, now that 1.1.0 is released, per the JIRA TO-DO. That's pretty much all there is to this.

Author: Sean Owen <sowen@cloudera.com>

Closes #3842 from srowen/SPARK-2757 and squashes the following commits:

50ff80e [Sean Owen] Exclude apparent false positive turned up by re-enabling MiMa checks for Streaming Flume Sink
0e5ba5c [Sean Owen] Re-enable MiMa for Streaming Flume Sink module

project/MimaExcludes.scala

@@ -53,6 +53,11 @@ object MimaExcludes {
               "org.apache.spark.mllib.linalg.Matrices.randn"),
             ProblemFilters.exclude[MissingMethodProblem](
               "org.apache.spark.mllib.linalg.Matrices.rand")
+          ) ++ Seq(
+            // SPARK-2757
+            ProblemFilters.exclude[IncompatibleResultTypeProblem](
+              "org.apache.spark.streaming.flume.sink.SparkAvroCallbackHandler." +
+                "removeAndGetProcessor")
           )
 
         case v if v.startsWith("1.2") =>

project/SparkBuild.scala

@@ -166,7 +166,7 @@ object SparkBuild extends PomBuild {
 
   // TODO: Add Sql to mima checks
   allProjects.filterNot(x => Seq(spark, sql, hive, hiveThriftServer, catalyst, repl,
-    streamingFlumeSink, networkCommon, networkShuffle, networkYarn).contains(x)).foreach {
+    networkCommon, networkShuffle, networkYarn).contains(x)).foreach {
       x => enable(MimaBuild.mimaSettings(sparkHome, x))(x)
     }