SPARK-3782 [CORE] Direct use of log4j in AkkaUtils interferes with certain logging configurations

Although the underlying issue can I think be solved by having user code use slf4j 1.7.6+, it might be helpful and consistent to update Spark's slf4j too. I see no reason to believe it would be incompatible with other 1.7.x releases: http://www.slf4j.org/news.html  Lots of different version of slf4j are in use in the wild and anecdotally I have never seen an issue mixing them.

Author: Sean Owen <sowen@cloudera.com>

Closes #4184 from srowen/SPARK-3782 and squashes the following commits:

5608d28 [Sean Owen] Update slf4j to 1.7.10

pom.xml

@@ -117,7 +117,7 @@
     <scala.macros.version>2.0.1</scala.macros.version>
     <mesos.version>0.21.0</mesos.version>
     <mesos.classifier>shaded-protobuf</mesos.classifier>
-    <slf4j.version>1.7.5</slf4j.version>
+    <slf4j.version>1.7.10</slf4j.version>
     <log4j.version>1.2.17</log4j.version>
     <hadoop.version>1.0.4</hadoop.version>
     <protobuf.version>2.4.1</protobuf.version>