Updated demo

README.md

@@ -116,4 +116,4 @@ Next, run Apache Spark SQL queries with Opaque as follows, assuming Spark is alr
 
 ## Contact
 
-If you want to know more about our project or have questions, please contact Ankur ([email protected]) and/or Wenting ([email protected]).
+If you want to know more about our project or have questions, please contact Wenting ([email protected]) and/or Ankur ([email protected]).

demo/demo.md

@@ -0,0 +1,56 @@
+# Opaque demo
+
+## Setup
+
+Follow the setup instructions in README (steps 1 - 3) to launch a Spark shell with Opaque. In order to try out our attack, you should also include `--conf "spark.driver.extraJavaOptions=-Xdebug -agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=8001"` when you launch Spark shell.
+
+## Usage example
+
+### Data creation
+
+```scala
+val data = for (i <- 0 until 10) yield ("foo", i)
+val rdd_data = spark.sparkContext.makeRDD(data, i)
+```
+
+### DataFrame creation
+```
+val words = spark.createDataFrame(rdd_data).toDF("word", "count")
+val words_e = spark.createDataFrame(rdd_data).toDF("word", "count").encrypted
+val words_o = spark.createDataFrame(rdd_data).toDF("word", "count").oblivious
+```
+
+### Query execution
+
+```
+words.filter($"count" > lit(3)).collect
+words_e.filter($"count" > lit(3)).collect
+words_o.filter($"count" > lit(3)).collect
+```
+
+## Attack example
+
+### Attach jdb
+`jdb -sourcepath $HOME/spark/sql/core/src/main/scala:$HOME/opaque/src/main/scala -attach 8000`
+
+### Attack Spark SQL
+
+`stop at org.apache.spark.sql.execution.FilterExec$$anonfun$12$$anonfun$apply$2:126
+list
+print row
+cont
+cont
+cont
+cont
+print row
+set r = false
+clear org.apache.spark.sql.execution.FilterExec$$anonfun$12$$anonfun$apply$2:126
+`
+
+### Attack Opaque (works for both encryption and oblivious modes)
+`stop at edu.berkeley.cs.rise.opaque.execution.ObliviousFilterExec$$anonfun$executeBlocked$14:323
+list
+dump filtered
+set filtered[0] = 123
+cont
+`

demo/opaque-demo.org

@@ -0,0 +1,52 @@
+* Opaque demo
+** Setup
+cd ~/spark
+source conf/spark-env.sh
+build/sbt package
+cd ~/opaque
+bin/spark-shell --master local[1] --jars $HOME/opaque/target/scala-2.11/opaque_2.11-0.1.jar --conf "spark.driver.extraJavaOptions=-Xdebug -agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=8001"
+
+** In Spark shell
+*** Import 
+import edu.berkeley.cs.rise.opaque.implicits._
+edu.berkeley.cs.rise.opaque.Utils.initSQLContext(spark.sqlContext)
+
+*** Data creation
+val data = for (i <- 0 until 10) yield ("foo", i)
+val rdd_data = spark.sparkContext.makeRDD(data, 1)
+
+*** DataFrame creation
+
+val words = spark.createDataFrame(rdd_data).toDF("word", "count")
+val words_e = spark.createDataFrame(rdd_data).toDF("word", "count").encrypted
+val words_o = spark.createDataFrame(rdd_data).toDF("word", "count").oblivious
+
+*** Query execution
+
+words.filter($"count" > lit(3)).collect
+words_e.filter($"count" > lit(3)).collect
+words_o.filter($"count" > lit(3)).collect
+
+** In JDB
+PS1="\[\033[01;38;5;210m\]\u@\h (attacker"'!'")\[\033[00m\] \D{%Y-%m-%d %H:%M:%S} \[\033[01;34m\]\w\[\033[00m\]\n\$ "
+
+jdb -sourcepath $HOME/spark/sql/core/src/main/scala:$HOME/opaque/src/main/scala -attach 8000
+
+*** For attacking insecure Spark
+stop at org.apache.spark.sql.execution.FilterExec$$anonfun$12$$anonfun$apply$2:126
+list
+print row
+cont
+cont
+cont
+cont
+print row
+set r = false
+clear org.apache.spark.sql.execution.FilterExec$$anonfun$12$$anonfun$apply$2:126
+
+*** For attempting to attack Opaque
+stop at edu.berkeley.cs.rise.opaque.execution.ObliviousFilterExec$$anonfun$executeBlocked$14:323
+list
+dump filtered
+set filtered[0] = 123
+cont