geode-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From jasonhu...@apache.org
Subject [07/10] incubator-geode git commit: GEODE-1244: Revert rename of package, directory, project and file rename for geode-spark-connector
Date Wed, 20 Apr 2016 17:33:41 GMT
http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSQLContextFunctions.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSQLContextFunctions.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSQLContextFunctions.scala
new file mode 100644
index 0000000..51f1d01
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSQLContextFunctions.scala
@@ -0,0 +1,42 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector
+
+import io.pivotal.gemfire.spark.connector.internal.oql.{OQLRelation, QueryRDD}
+import org.apache.spark.Logging
+import org.apache.spark.sql.{DataFrame, SQLContext}
+
+/**
+ * Provide GemFire OQL specific functions
+ */
+class GemFireSQLContextFunctions(@transient sqlContext: SQLContext) extends Serializable with Logging {
+
+  /**
+   * Expose a GemFire OQL query result as a DataFrame
+   * @param query the OQL query string.
+   */
+  def gemfireOQL(
+    query: String,
+    connConf: GemFireConnectionConf = GemFireConnectionConf(sqlContext.sparkContext.getConf)): DataFrame = {
+    logInfo(s"OQL query = $query")
+    val rdd = new QueryRDD[Object](sqlContext.sparkContext, query, connConf)
+    sqlContext.baseRelationToDataFrame(OQLRelation(rdd)(sqlContext))
+  }
+
+  private[connector] def defaultConnectionConf: GemFireConnectionConf =
+    GemFireConnectionConf(sqlContext.sparkContext.getConf)
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSparkContextFunctions.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSparkContextFunctions.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSparkContextFunctions.scala
new file mode 100644
index 0000000..5341977
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/GemFireSparkContextFunctions.scala
@@ -0,0 +1,39 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector
+
+import io.pivotal.gemfire.spark.connector.internal.rdd.GemFireRegionRDD
+import org.apache.spark.SparkContext
+
+import scala.reflect.ClassTag
+
+/** Provides GemFire specific methods on `SparkContext` */
+class GemFireSparkContextFunctions(@transient sc: SparkContext) extends Serializable {
+
+  /**
+   * Expose a GemFire region as a GemFireRDD
+   * @param regionPath the full path of the region
+   * @param connConf the GemFireConnectionConf that can be used to access the region
+   * @param opConf use this to specify preferred partitioner
+   *        and its parameters. The implementation will use it if it's applicable
+   */
+  def gemfireRegion[K: ClassTag, V: ClassTag] (
+    regionPath: String, connConf: GemFireConnectionConf = GemFireConnectionConf(sc.getConf),
+    opConf: Map[String, String] = Map.empty): GemFireRegionRDD[K, V] =
+    GemFireRegionRDD[K, V](sc, regionPath, connConf, opConf)
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnection.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnection.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnection.scala
new file mode 100644
index 0000000..7d147b2
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnection.scala
@@ -0,0 +1,164 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal
+
+import java.net.InetAddress
+
+import com.gemstone.gemfire.cache.client.{ClientCache, ClientCacheFactory, ClientRegionShortcut}
+import com.gemstone.gemfire.cache.execute.{FunctionException, FunctionService}
+import com.gemstone.gemfire.cache.query.Query
+import com.gemstone.gemfire.cache.{Region, RegionService}
+import com.gemstone.gemfire.internal.cache.execute.InternalExecution
+import io.pivotal.gemfire.spark.connector.internal.oql.QueryResultCollector
+import io.pivotal.gemfire.spark.connector.internal.rdd.GemFireRDDPartition
+import org.apache.spark.{SparkEnv, Logging}
+import io.pivotal.gemfire.spark.connector.GemFireConnection
+import io.pivotal.gemfire.spark.connector.internal.gemfirefunctions._
+import java.util.{Set => JSet, List => JList }
+
+/**
+ * Default GemFireConnection implementation. The instance of this should be
+ * created by DefaultGemFireConnectionFactory
+ * @param locators pairs of host/port of locators
+ * @param gemFireProps The initial gemfire properties to be used.
+ */
+private[connector] class DefaultGemFireConnection (
+  locators: Seq[(String, Int)], gemFireProps: Map[String, String] = Map.empty) 
+  extends GemFireConnection with Logging {
+
+  private val clientCache = initClientCache()
+
+  /** Register GemFire functions to the GemFire cluster */
+  FunctionService.registerFunction(RetrieveRegionMetadataFunction.getInstance())
+  FunctionService.registerFunction(RetrieveRegionFunction.getInstance())
+
+  private def initClientCache() : ClientCache = {
+    try {
+      val ccf = getClientCacheFactory
+      ccf.create()
+    } catch {
+      case e: Exception =>
+        logError(s"""Failed to init ClientCache, locators=${locators.mkString(",")}, Error: $e""")
+        throw new RuntimeException(e)
+    }
+  }
+  
+  private def getClientCacheFactory: ClientCacheFactory = {
+    import io.pivotal.gemfire.spark.connector.map2Properties
+    val ccf = new ClientCacheFactory(gemFireProps)
+    ccf.setPoolReadTimeout(30000)
+    val servers = LocatorHelper.getAllGemFireServers(locators)
+    if (servers.isDefined && servers.get.size > 0) {
+      val sparkIp = System.getenv("SPARK_LOCAL_IP")
+      val hostName = if (sparkIp != null) InetAddress.getByName(sparkIp).getCanonicalHostName
+                     else InetAddress.getLocalHost.getCanonicalHostName
+      val executorId = SparkEnv.get.executorId      
+      val pickedServers = LocatorHelper.pickPreferredGemFireServers(servers.get, hostName, executorId)
+      logInfo(s"""Init ClientCache: severs=${pickedServers.mkString(",")}, host=$hostName executor=$executorId props=$gemFireProps""")
+      logDebug(s"""Init ClientCache: all-severs=${pickedServers.mkString(",")}""")
+      pickedServers.foreach{ case (host, port)  => ccf.addPoolServer(host, port) }
+    } else {
+      logInfo(s"""Init ClientCache: locators=${locators.mkString(",")}, props=$gemFireProps""")
+      locators.foreach { case (host, port)  => ccf.addPoolLocator(host, port) }
+    }
+    ccf
+  }
+
+  /** close the clientCache */
+  override def close(): Unit =
+    if (! clientCache.isClosed) clientCache.close()
+
+  /** ----------------------------------------- */
+  /** implementation of GemFireConnection trait */
+  /** ----------------------------------------- */
+
+  override def getQuery(queryString: String): Query =
+    clientCache.asInstanceOf[RegionService].getQueryService.newQuery(queryString)
+
+  override def validateRegion[K, V](regionPath: String): Unit = {
+    val md = getRegionMetadata[K, V](regionPath)
+    if (! md.isDefined) throw new RuntimeException(s"The region named $regionPath was not found")
+  }
+
+  def getRegionMetadata[K, V](regionPath: String): Option[RegionMetadata] = {
+    import scala.collection.JavaConversions.setAsJavaSet
+    val region = getRegionProxy[K, V](regionPath)
+    val set0: JSet[Integer] = Set[Integer](0)
+    val exec = FunctionService.onRegion(region).asInstanceOf[InternalExecution].withBucketFilter(set0)
+    exec.setWaitOnExceptionFlag(true)
+    try {
+      val collector = exec.execute(RetrieveRegionMetadataFunction.ID)
+      val r = collector.getResult.asInstanceOf[JList[RegionMetadata]]
+      logDebug(r.get(0).toString)
+      Some(r.get(0))
+    } catch {
+      case e: FunctionException => 
+        if (e.getMessage.contains(s"The region named /$regionPath was not found")) None
+        else throw e
+    }
+  }
+
+  def getRegionProxy[K, V](regionPath: String): Region[K, V] = {
+    val region1: Region[K, V] = clientCache.getRegion(regionPath).asInstanceOf[Region[K, V]]
+    if (region1 != null) region1
+    else DefaultGemFireConnection.regionLock.synchronized {
+      val region2 = clientCache.getRegion(regionPath).asInstanceOf[Region[K, V]]
+      if (region2 != null) region2
+      else clientCache.createClientRegionFactory[K, V](ClientRegionShortcut.PROXY).create(regionPath)
+    }
+  }
+
+  override def getRegionData[K, V](regionPath: String, whereClause: Option[String], split: GemFireRDDPartition): Iterator[(K, V)] = {
+    val region = getRegionProxy[K, V](regionPath)
+    val desc = s"""RDD($regionPath, "${whereClause.getOrElse("")}", ${split.index})"""
+    val args : Array[String] = Array[String](whereClause.getOrElse(""), desc)
+    val collector = new StructStreamingResultCollector(desc)
+        // RetrieveRegionResultCollector[(K, V)]
+    import scala.collection.JavaConversions.setAsJavaSet
+    val exec = FunctionService.onRegion(region).withArgs(args).withCollector(collector).asInstanceOf[InternalExecution]
+      .withBucketFilter(split.bucketSet.map(Integer.valueOf))
+    exec.setWaitOnExceptionFlag(true)
+    exec.execute(RetrieveRegionFunction.ID)
+    collector.getResult.map{objs: Array[Object] => (objs(0).asInstanceOf[K], objs(1).asInstanceOf[V])}
+  }
+
+  override def executeQuery(regionPath: String, bucketSet: Set[Int], queryString: String) = {
+    import scala.collection.JavaConversions.setAsJavaSet
+    FunctionService.registerFunction(QueryFunction.getInstance())
+    val collector = new QueryResultCollector
+    val region = getRegionProxy(regionPath)
+    val args: Array[String] = Array[String](queryString, bucketSet.toString)
+    val exec = FunctionService.onRegion(region).withCollector(collector).asInstanceOf[InternalExecution]
+      .withBucketFilter(bucketSet.map(Integer.valueOf))
+      .withArgs(args)
+    exec.execute(QueryFunction.ID)
+    collector.getResult
+  }
+}
+
+private[connector] object DefaultGemFireConnection {
+  /** a lock object only used by getRegionProxy...() */
+  private val regionLock = new Object
+}
+
+/** The purpose of this class is making unit test DefaultGemFireConnectionManager easier */
+class DefaultGemFireConnectionFactory {
+
+  def newConnection(locators: Seq[(String, Int)], gemFireProps: Map[String, String] = Map.empty) =
+    new DefaultGemFireConnection(locators, gemFireProps)
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnectionManager.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnectionManager.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnectionManager.scala
new file mode 100644
index 0000000..6722ca8
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/DefaultGemFireConnectionManager.scala
@@ -0,0 +1,77 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal
+
+import io.pivotal.gemfire.spark.connector.{GemFireConnection, GemFireConnectionConf, GemFireConnectionManager}
+
+import scala.collection.mutable
+
+/**
+ * Default implementation of GemFireConnectionFactory
+ */
+class DefaultGemFireConnectionManager extends GemFireConnectionManager {
+
+  def getConnection(connConf: GemFireConnectionConf): GemFireConnection =
+    DefaultGemFireConnectionManager.getConnection(connConf)
+
+  def closeConnection(connConf: GemFireConnectionConf): Unit =
+    DefaultGemFireConnectionManager.closeConnection(connConf)
+
+}
+
+object DefaultGemFireConnectionManager  {
+
+  /** connection cache, keyed by host:port pair */
+  private[connector] val connections = mutable.Map[(String, Int), GemFireConnection]()
+
+  /**
+   * use locator host:port pair to lookup cached connection. create new connection 
+   * and add it to the cache `connections` if it does not exist.
+   */
+  def getConnection(connConf: GemFireConnectionConf)
+    (implicit factory: DefaultGemFireConnectionFactory = new DefaultGemFireConnectionFactory): GemFireConnection = {
+
+    def getCachedConnection(locators: Seq[(String, Int)]): GemFireConnection = {
+      val conns = connConf.locators.map(connections withDefaultValue null).filter(_ != null)
+      if (conns.nonEmpty) conns(0) else null
+    }
+
+    val conn1 = getCachedConnection(connConf.locators)
+    if (conn1 != null) conn1
+    else connections.synchronized {
+      val conn2 = getCachedConnection(connConf.locators)
+      if (conn2 != null) conn2
+      else {
+        val conn3 = factory.newConnection(connConf.locators, connConf.gemfireProps)
+        connConf.locators.foreach(pair => connections += (pair -> conn3))
+        conn3
+      }
+    }
+  }
+
+  /**
+   * Close the connection and remove it from connection cache.
+   * Note: multiple entries may share the same connection, all those entries are removed.
+   */
+  def closeConnection(connConf: GemFireConnectionConf): Unit = {
+    val conns = connConf.locators.map(connections withDefaultValue null).filter(_ != null)
+    if (conns.nonEmpty) connections.synchronized {
+      conns(0).close()
+      connections.retain((k,v) => v != conns(0))
+    }
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/LocatorHelper.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/LocatorHelper.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/LocatorHelper.scala
new file mode 100644
index 0000000..1d72775
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/LocatorHelper.scala
@@ -0,0 +1,135 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal
+
+import java.net.InetSocketAddress
+import java.util.{ArrayList => JArrayList}
+
+import com.gemstone.gemfire.cache.client.internal.locator.{GetAllServersResponse, GetAllServersRequest}
+import com.gemstone.gemfire.distributed.internal.ServerLocation
+import com.gemstone.gemfire.distributed.internal.tcpserver.TcpClient
+import org.apache.spark.Logging
+
+import scala.util.{Failure, Success, Try}
+
+
+object LocatorHelper extends Logging {
+
+  /** valid locator strings are: host[port] and host:port */
+  final val LocatorPattern1 = """([\w-_]+(\.[\w-_]+)*)\[([0-9]{2,5})\]""".r
+  final val LocatorPattern2 = """([\w-_]+(\.[\w-_]+)*):([0-9]{2,5})""".r
+
+  /** convert single locator string to Try[(host, port)] */
+  def locatorStr2HostPortPair(locatorStr: String): Try[(String, Int)] =
+    locatorStr match {
+      case LocatorPattern1(host, domain, port) => Success((host, port.toInt))
+      case LocatorPattern2(host, domain, port) => Success((host, port.toInt))
+      case _ => Failure(new Exception(s"invalid locator: $locatorStr"))
+    }
+
+  /** 
+   * Parse locator strings and returns Seq of (hostname, port) pair. 
+   * Valid locator string are one or more "host[port]" and/or "host:port"
+   * separated by `,`. For example:
+   *    host1.mydomain.com[8888],host2.mydomain.com[8889] 
+   *    host1.mydomain.com:8888,host2.mydomain.com:8889 
+   */
+  def parseLocatorsString(locatorsStr: String): Seq[(String, Int)] =
+    locatorsStr.split(",").map(locatorStr2HostPortPair).map(_.get)
+
+
+  /**
+   * Return the list of live GemFire servers for the given locators. 
+   * @param locators locators for the given GemFire cluster
+   * @param serverGroup optional server group name, default is "" (empty string)
+   */
+  def getAllGemFireServers(locators: Seq[(String, Int)], serverGroup: String = ""): Option[Seq[(String, Int)]] = {
+    var result: Option[Seq[(String, Int)]] = None
+    locators.find { case (host, port) =>
+      try {
+        val addr = new InetSocketAddress(host, port)
+        val req = new GetAllServersRequest(serverGroup)
+        val res = TcpClient.requestToServer(addr.getAddress, addr.getPort, req, 2000)
+        if (res != null) {
+          import scala.collection.JavaConverters._
+          val servers = res.asInstanceOf[GetAllServersResponse].getServers.asInstanceOf[JArrayList[ServerLocation]]
+          if (servers.size > 0)
+            result = Some(servers.asScala.map(e => (e.getHostName, e.getPort)))
+        }
+      } catch { case e: Exception => logWarning("getAllGemFireServers error", e)
+      }
+      result.isDefined
+    }
+    result
+  }
+
+  /**
+   * Pick up at most 3 preferred servers from all available servers based on
+   * host name and Spark executor id.
+   *
+   * This method is used by DefaultGemFireConnection to create ClientCache. Usually
+   * one server is enough to initialize ClientCacheFactory, but this provides two
+   * backup servers in case of the 1st server can't be connected.
+   *   
+   * @param servers all available servers in the form of (hostname, port) pairs
+   * @param hostName the host name of the Spark executor
+   * @param executorId the Spark executor Id, such as "<driver>", "0", "1", ...
+   * @return Seq[(hostname, port)] of preferred servers
+   */
+  def pickPreferredGemFireServers(
+    servers: Seq[(String, Int)], hostName: String, executorId: String): Seq[(String, Int)] = {
+
+    // pick up `length` items form the Seq starts at the `start` position.
+    //  The Seq is treated as a ring, so at most `Seq.size` items can be picked
+    def circularTake[T](seq: Seq[T], start: Int, length: Int): Seq[T] = {
+      val size = math.min(seq.size, length)
+      (start until start + size).map(x => seq(x % seq.size))
+    }
+
+    // map executor id to int: "<driver>" (or non-number string) to 0, and "n" to n + 1
+    val id = try { executorId.toInt + 1 } catch { case e: NumberFormatException => 0 }
+    
+    // algorithm: 
+    // 1. sort server list
+    // 2. split sorted server list into 3 sub-lists a, b, and c:
+    //      list-a: servers on the given host
+    //      list-b: servers that are in front of list-a on the sorted server list
+    //      list-c: servers that are behind list-a on the sorted server list
+    //    then rotate list-a based on executor id, then create new server list:
+    //      modified list-a ++ list-c ++ list-b
+    // 3. if there's no server on the given host, then create new server list
+    //    by rotating sorted server list based on executor id.
+    // 4. take up to 3 servers from the new server list
+    val sortedServers = servers.sorted
+    val firstIdx = sortedServers.indexWhere(p => p._1 == hostName)
+    val lastIdx = if (firstIdx < 0) -1 else sortedServers.lastIndexWhere(p => p._1 == hostName)
+
+    if (firstIdx < 0) { // no local server
+      circularTake(sortedServers, id, 3)
+    } else {
+      val (seq1, seq2) = sortedServers.splitAt(firstIdx)
+      val seq = if (firstIdx == lastIdx) {  // one local server
+        seq2 ++ seq1
+      } else { // multiple local server
+        val (seq3, seq4) = seq2.splitAt(lastIdx - firstIdx + 1)
+        val seq3b = if (id % seq3.size == 0) seq3 else circularTake(seq3, id, seq3.size)
+        seq3b ++ seq4 ++ seq1
+      }
+      circularTake(seq, 0, 3)
+    }
+  }  
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/gemfirefunctions/StructStreamingResultCollector.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/gemfirefunctions/StructStreamingResultCollector.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/gemfirefunctions/StructStreamingResultCollector.scala
new file mode 100644
index 0000000..39bc0cc
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/gemfirefunctions/StructStreamingResultCollector.scala
@@ -0,0 +1,152 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.gemfirefunctions
+
+import java.util.concurrent.{TimeUnit, LinkedBlockingQueue, BlockingQueue}
+import com.gemstone.gemfire.DataSerializer
+import com.gemstone.gemfire.cache.execute.ResultCollector
+import com.gemstone.gemfire.cache.query.internal.types.StructTypeImpl
+import com.gemstone.gemfire.cache.query.types.StructType
+import com.gemstone.gemfire.distributed.DistributedMember
+import com.gemstone.gemfire.internal.{Version, ByteArrayDataInput}
+import io.pivotal.gemfire.spark.connector.internal.gemfirefunctions.StructStreamingResultSender.
+       {TYPE_CHUNK, DATA_CHUNK, ERROR_CHUNK, SER_DATA, UNSER_DATA, BYTEARR_DATA}
+
+/**
+ * StructStreamingResultCollector and StructStreamingResultSender are paired
+ * to transfer result of list of `com.gemstone.gemfire.cache.query.Struct`
+ * from GemFire server to Spark Connector (the client of GemFire server) 
+ * in streaming, i.e., while sender sending the result, the collector can
+ * start processing the arrived result without waiting for full result to
+ * become available.
+ */
+class StructStreamingResultCollector(desc: String) extends ResultCollector[Array[Byte], Iterator[Array[Object]]] {
+
+  /** the constructor that provide default `desc` (description) */
+  def this() = this("StructStreamingResultCollector")
+  
+  private val queue: BlockingQueue[Array[Byte]] = new LinkedBlockingQueue[Array[Byte]]()
+  var structType: StructType = null
+
+  /** ------------------------------------------ */
+  /**  ResultCollector interface implementations */
+  /** ------------------------------------------ */
+  
+  override def getResult: Iterator[Array[Object]] = resultIterator
+
+  override def getResult(timeout: Long, unit: TimeUnit): Iterator[Array[Object]] = 
+    throw new UnsupportedOperationException()
+
+  /** addResult add non-empty byte array (chunk) to the queue */
+  override def addResult(memberID: DistributedMember, chunk: Array[Byte]): Unit = 
+    if (chunk != null && chunk.size > 1) {
+      this.queue.add(chunk)
+      // println(s"""$desc receive from $memberID: ${chunk.mkString(" ")}""")
+    }
+
+  /** endResults add special `Array.empty` to the queue as marker of end of data */
+  override def endResults(): Unit = this.queue.add(Array.empty)
+  
+  override def clearResults(): Unit = this.queue.clear()
+
+  /** ------------------------------------------ */
+  /**             Internal methods               */
+  /** ------------------------------------------ */
+
+  def getResultType: StructType = {
+    // trigger lazy resultIterator initialization if necessary
+    if (structType == null)  resultIterator.hasNext
+    structType        
+  }
+
+  /**
+   * Note: The data is sent in chunks, and each chunk contains multiple 
+   * records. So the result iterator is an iterator (I) of iterator (II),
+   * i.e., go through each chunk (iterator (I)), and for each chunk, go 
+   * through each record (iterator (II)). 
+   */
+  private lazy val resultIterator = new Iterator[Array[Object]] {
+
+    private var currentIterator: Iterator[Array[Object]] = nextIterator()
+    
+    override def hasNext: Boolean = {
+      if (!currentIterator.hasNext && currentIterator != Iterator.empty) currentIterator = nextIterator()
+      currentIterator.hasNext
+    }
+
+    /** Note: make sure call `hasNext` first to adjust `currentIterator` */
+    override def next(): Array[Object] = currentIterator.next()
+  }
+  
+  /** get the iterator for the next chunk of data */
+  private def nextIterator(): Iterator[Array[Object]] = {
+    val chunk: Array[Byte] = queue.take
+    if (chunk.isEmpty) {
+      Iterator.empty
+    } else {
+      val input = new ByteArrayDataInput()
+      input.initialize(chunk, Version.CURRENT)
+      val chunkType = input.readByte()
+      // println(s"chunk type $chunkType")
+      chunkType match {
+        case TYPE_CHUNK =>
+          if (structType == null)
+            structType = DataSerializer.readObject(input).asInstanceOf[StructTypeImpl]
+          nextIterator()
+        case DATA_CHUNK =>
+          // require(structType != null && structType.getFieldNames.length > 0)
+          if (structType == null) structType = StructStreamingResultSender.KeyValueType
+          chunkToIterator(input, structType.getFieldNames.length)
+        case ERROR_CHUNK => 
+          val error = DataSerializer.readObject(input).asInstanceOf[Exception]
+          errorPropagationIterator(error)
+        case _ => throw new RuntimeException(s"unknown chunk type: $chunkType")
+      }
+    }
+  }
+
+  /** create a iterator that propagate sender's exception */
+  private def errorPropagationIterator(ex: Exception) = new Iterator[Array[Object]] {
+    val re = new RuntimeException(ex)
+    override def hasNext: Boolean = throw re
+    override def next(): Array[Object] = throw re
+  }
+  
+  /** convert a chunk of data to an iterator */
+  private def chunkToIterator(input: ByteArrayDataInput, rowSize: Int) = new Iterator[Array[Object]] {
+    override def hasNext: Boolean = input.available() > 0
+    val tmpInput = new ByteArrayDataInput()
+    override def next(): Array[Object] = 
+      (0 until rowSize).map { ignore =>
+        val b = input.readByte()
+        b match {
+          case SER_DATA => 
+            val arr: Array[Byte] = DataSerializer.readByteArray(input)
+            tmpInput.initialize(arr, Version.CURRENT)
+            DataSerializer.readObject(tmpInput).asInstanceOf[Object]
+          case UNSER_DATA =>
+            DataSerializer.readObject(input).asInstanceOf[Object]
+          case BYTEARR_DATA =>
+            DataSerializer.readByteArray(input).asInstanceOf[Object]
+          case _ => 
+            throw new RuntimeException(s"unknown data type $b")
+        }
+      }.toArray
+  }
+  
+}
+

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryParser.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryParser.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryParser.scala
new file mode 100644
index 0000000..71e1823
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryParser.scala
@@ -0,0 +1,58 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import scala.util.parsing.combinator.RegexParsers
+
+class QueryParser extends RegexParsers {
+
+  def query: Parser[String] = opt(rep(IMPORT ~ PACKAGE)) ~> select ~> opt(distinct) ~> projection ~> from ~> regions <~ opt(where ~ filter) ^^ {
+    _.toString
+  }
+
+  val IMPORT: Parser[String] = "[Ii][Mm][Pp][Oo][Rr][Tt]".r
+
+  val select: Parser[String] = "[Ss][Ee][Ll][Ee][Cc][Tt]".r
+
+  val distinct: Parser[String] = "[Dd][Ii][Ss][Tt][Ii][Nn][Cc][Tt]".r
+
+  val from: Parser[String] = "[Ff][Rr][Oo][Mm]".r
+
+  val where: Parser[String] = "[Ww][Hh][Ee][Rr][Ee]".r
+
+  def PACKAGE: Parser[String] = """[\w.]+""".r
+
+  def projection: Parser[String] = "*" | repsep("""["\w]+[.\w"]*""".r, ",") ^^ {
+    _.toString
+  }
+
+  def regions: Parser[String] = repsep(region <~ opt(alias), ",") ^^ {
+    _.toString
+  }
+
+  def region: Parser[String] = """/[\w.]+[/[\w.]+]*""".r | """[\w]+[.\w]*""".r
+
+  def alias: Parser[String] = not(where) ~> """[\w]+""".r
+
+  def filter: Parser[String] = """[\w.]+[[\s]+[<>=.'\w]+]*""".r
+}
+
+object QueryParser extends QueryParser {
+
+  def parseOQL(expression: String) = parseAll(query, expression)
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryRDD.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryRDD.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryRDD.scala
new file mode 100644
index 0000000..c16a70d
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryRDD.scala
@@ -0,0 +1,83 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import io.pivotal.gemfire.spark.connector.GemFireConnectionConf
+import io.pivotal.gemfire.spark.connector.internal.rdd.{GemFireRDDPartition, ServerSplitsPartitioner}
+import org.apache.spark.rdd.RDD
+import org.apache.spark.{TaskContext, SparkContext, Partition}
+import scala.reflect.ClassTag
+
+/**
+ * An RDD that provides the functionality that read the OQL query result
+ *
+ * @param sc The SparkContext this RDD is associated with
+ * @param queryString The OQL query string
+ * @param connConf The GemFireConnectionConf that provide the GemFireConnection
+ */
+class QueryRDD[T](@transient sc: SparkContext,
+                  queryString: String,
+                  connConf: GemFireConnectionConf)
+                 (implicit ct: ClassTag[T])
+  extends RDD[T](sc, Seq.empty) {
+
+  override def getPartitions: Array[Partition] = {
+    val conn = connConf.getConnection
+    val regionPath = getRegionPathFromQuery(queryString)
+    val md = conn.getRegionMetadata(regionPath)
+    md match {
+      case Some(metadata) =>
+        if (metadata.isPartitioned) {
+          val splits = ServerSplitsPartitioner.partitions(conn, metadata, Map.empty)
+          logInfo(s"QueryRDD.getPartitions():isPartitioned=true, partitions=${splits.mkString(",")}")
+          splits
+        }
+        else {
+          logInfo(s"QueryRDD.getPartitions():isPartitioned=false")
+          Array[Partition](new GemFireRDDPartition(0, Set.empty))
+
+        }
+      case None => throw new RuntimeException(s"Region $regionPath metadata was not found.")
+    }
+  }
+
+  override def compute(split: Partition, context: TaskContext): Iterator[T] = {
+    val buckets = split.asInstanceOf[GemFireRDDPartition].bucketSet
+    val regionPath = getRegionPathFromQuery(queryString)
+    val result = connConf.getConnection.executeQuery(regionPath, buckets, queryString)
+    result match {
+      case it: Iterator[T] =>
+        logInfo(s"QueryRDD.compute():query=$queryString, partition=$split")
+        it
+      case _ =>
+        throw new RuntimeException("Unexpected OQL result: " + result.toString)
+    }
+  }
+
+  private def getRegionPathFromQuery(queryString: String): String = {
+    val r = QueryParser.parseOQL(queryString).get
+    r match {
+      case r: String =>
+        val start = r.indexOf("/") + 1
+        var end = r.indexOf(")")
+        if (r.indexOf(".") > 0) end = math.min(r.indexOf("."), end)
+        if (r.indexOf(",") > 0) end = math.min(r.indexOf(","), end)
+        val regionPath = r.substring(start, end)
+        regionPath
+    }
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryResultCollector.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryResultCollector.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryResultCollector.scala
new file mode 100644
index 0000000..7032e5a
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/QueryResultCollector.scala
@@ -0,0 +1,69 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import java.util.concurrent.{TimeUnit, LinkedBlockingDeque}
+
+import com.gemstone.gemfire.DataSerializer
+import com.gemstone.gemfire.cache.execute.ResultCollector
+import com.gemstone.gemfire.distributed.DistributedMember
+import com.gemstone.gemfire.internal.{Version, ByteArrayDataInput}
+
+class QueryResultCollector extends ResultCollector[Array[Byte], Iterator[Object]]{
+
+  private val queue = new LinkedBlockingDeque[Array[Byte]]()
+
+  override def getResult = resultIterator
+
+  override def getResult(timeout: Long, unit: TimeUnit) = throw new UnsupportedOperationException
+
+  override def addResult(memberID: DistributedMember , chunk: Array[Byte]) =
+    if (chunk != null && chunk.size > 0) {
+      queue.add(chunk)
+    }
+
+  override def endResults = queue.add(Array.empty)
+
+
+  override def clearResults = queue.clear
+
+  private lazy val resultIterator = new Iterator[Object] {
+    private var currentIterator = nextIterator
+    def hasNext = {
+      if (!currentIterator.hasNext && currentIterator != Iterator.empty)
+        currentIterator = nextIterator
+      currentIterator.hasNext
+    }
+    def next = currentIterator.next
+  }
+
+  private def nextIterator: Iterator[Object] = {
+    val chunk = queue.take
+    if (chunk.isEmpty) {
+      Iterator.empty
+    }
+    else {
+      val input = new ByteArrayDataInput
+      input.initialize(chunk, Version.CURRENT)
+      new Iterator[Object] {
+        override def hasNext: Boolean = input.available() > 0
+        override def next: Object = DataSerializer.readObject(input).asInstanceOf[Object]
+      }
+    }
+  }
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RDDConverter.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RDDConverter.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RDDConverter.scala
new file mode 100644
index 0000000..894066c
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RDDConverter.scala
@@ -0,0 +1,40 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import org.apache.spark.rdd.RDD
+import org.apache.spark.sql.{DataFrame, SQLContext}
+import org.apache.spark.sql.types._
+import org.apache.spark.sql.Row
+import org.apache.spark.sql.sources.{BaseRelation, TableScan}
+
+import scala.tools.nsc.backend.icode.analysis.DataFlowAnalysis
+
+case class OQLRelation[T](queryRDD: QueryRDD[T])(@transient val sqlContext: SQLContext) extends BaseRelation with TableScan {
+
+  override def schema: StructType = new SchemaBuilder(queryRDD).toSparkSchema()
+
+  override def buildScan(): RDD[Row] = new RowBuilder(queryRDD).toRowRDD()
+
+}
+
+object RDDConverter {
+
+  def queryRDDToDataFrame[T](queryRDD: QueryRDD[T], sqlContext: SQLContext): DataFrame = {
+    sqlContext.baseRelationToDataFrame(OQLRelation(queryRDD)(sqlContext))
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RowBuilder.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RowBuilder.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RowBuilder.scala
new file mode 100644
index 0000000..3a4fa6a
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/RowBuilder.scala
@@ -0,0 +1,38 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import com.gemstone.gemfire.cache.query.internal.StructImpl
+import org.apache.spark.rdd.RDD
+import org.apache.spark.sql._
+
+class RowBuilder[T](queryRDD: QueryRDD[T]) {
+
+  /**
+   * Convert QueryRDD to RDD of Row
+   * @return RDD of Rows
+   */
+  def toRowRDD(): RDD[Row] = {
+    val rowRDD = queryRDD.map(row => {
+      row match {
+        case si: StructImpl => Row.fromSeq(si.getFieldValues)
+        case obj: Object => Row.fromSeq(Seq(obj))
+      }
+    })
+    rowRDD
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/SchemaBuilder.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/SchemaBuilder.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/SchemaBuilder.scala
new file mode 100644
index 0000000..11fff90
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/SchemaBuilder.scala
@@ -0,0 +1,73 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import com.gemstone.gemfire.cache.query.internal.StructImpl
+import org.apache.spark.sql.types._
+import scala.collection.mutable.ListBuffer
+import org.apache.spark.Logging
+
+class SchemaBuilder[T](queryRDD: QueryRDD[T]) extends Logging {
+
+  val nullStructType = StructType(Nil)
+  
+  val typeMap:Map[Class[_], DataType] = Map( 
+    (classOf[java.lang.String], StringType),
+    (classOf[java.lang.Integer], IntegerType),
+    (classOf[java.lang.Short], ShortType),
+    (classOf[java.lang.Long], LongType),
+    (classOf[java.lang.Double], DoubleType),
+    (classOf[java.lang.Float], FloatType),
+    (classOf[java.lang.Boolean], BooleanType),
+    (classOf[java.lang.Byte], ByteType),
+    (classOf[java.util.Date], DateType),
+    (classOf[java.lang.Object], nullStructType)
+  )
+  
+  /**
+   * Analyse QueryRDD to get the Spark schema
+   * @return The schema represented by Spark StructType
+   */
+  def toSparkSchema(): StructType = {
+    val row = queryRDD.first()
+    val tpe = row match {
+      case r: StructImpl => constructFromStruct(r)
+      case null => StructType(StructField("col1", NullType) :: Nil)
+      case default => 
+        val value = typeMap.getOrElse(default.getClass(), nullStructType)
+        StructType(StructField("col1", value) :: Nil)
+    }
+    logInfo(s"Schema: $tpe")
+    tpe
+  }
+  
+  def constructFromStruct(r:StructImpl) = {
+    val names = r.getFieldNames
+    val values = r.getFieldValues
+    val lb = new ListBuffer[StructField]()
+    for (i <- 0 until names.length) {
+      val name = names(i)
+      val value = values(i)
+      val dataType = value match {
+        case null => NullType
+        case default => typeMap.getOrElse(default.getClass,  nullStructType)
+      }
+      lb += StructField(name, dataType)
+    }
+    StructType(lb.toSeq)
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/UndefinedSerializer.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/UndefinedSerializer.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/UndefinedSerializer.scala
new file mode 100644
index 0000000..35e8120
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/oql/UndefinedSerializer.scala
@@ -0,0 +1,46 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.oql
+
+import com.esotericsoftware.kryo.{Kryo, Serializer}
+import com.esotericsoftware.kryo.io.{Output, Input}
+import com.gemstone.gemfire.cache.query.QueryService
+import com.gemstone.gemfire.cache.query.internal.Undefined
+
+/**
+ * This is the customized serializer to serialize QueryService.UNDEFINED,
+ * i.e. com.gemstone.gemfire.cache.query.internal.Undefined, in order to
+ * guarantee the singleton Undefined after its deserialization within Spark.
+ */
+class UndefinedSerializer extends Serializer[Undefined] {
+
+  def write(kryo: Kryo, output: Output, u: Undefined) {
+    //Only serialize a byte for Undefined
+    output.writeByte(u.getDSFID)
+  }
+
+  def read (kryo: Kryo, input: Input, tpe: Class[Undefined]): Undefined = {
+    //Read DSFID of Undefined
+    input.readByte()
+    QueryService.UNDEFINED match {
+      case null => new Undefined
+      case _ =>
+        //Avoid calling Undefined constructor again.
+        QueryService.UNDEFINED.asInstanceOf[Undefined]
+    }
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireJoinRDD.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireJoinRDD.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireJoinRDD.scala
new file mode 100644
index 0000000..cfa1537
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireJoinRDD.scala
@@ -0,0 +1,67 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import com.gemstone.gemfire.cache.Region
+import io.pivotal.gemfire.spark.connector.GemFireConnectionConf
+import org.apache.spark.{TaskContext, Partition}
+import org.apache.spark.rdd.RDD
+import scala.collection.JavaConversions._
+
+/**
+ * An `RDD[T, V]` that will represent the result of a join between `left` RDD[T]
+ * and the specified GemFire Region[K, V].
+ */
+class GemFireJoinRDD[T, K, V] private[connector]
+  ( left: RDD[T],
+    func: T => K,
+    val regionPath: String,
+    val connConf: GemFireConnectionConf
+  ) extends RDD[(T, V)](left.context, left.dependencies) {
+
+  /** validate region existence when GemFireRDD object is created */
+  validate()
+
+  /** Validate region, and make sure it exists. */
+  private def validate(): Unit = connConf.getConnection.validateRegion[K, V](regionPath)
+
+  override protected def getPartitions: Array[Partition] = left.partitions
+
+  override def compute(split: Partition, context: TaskContext): Iterator[(T, V)] = {
+    val region = connConf.getConnection.getRegionProxy[K, V](regionPath)
+    if (func == null) computeWithoutFunc(split, context, region)
+    else computeWithFunc(split, context, region)
+  }
+
+  /** T is (K, V1) since there's no map function `func` */
+  private def computeWithoutFunc(split: Partition, context: TaskContext, region: Region[K, V]): Iterator[(T, V)] = {
+    val leftPairs = left.iterator(split, context).toList.asInstanceOf[List[(K, _)]]
+    val leftKeys = leftPairs.map { case (k, v) => k}.toSet
+    // Note: get all will return (key, null) for non-exist entry, so remove those entries
+    val rightPairs = region.getAll(leftKeys).filter { case (k, v) => v != null}
+    leftPairs.filter{case (k, v) => rightPairs.contains(k)}
+             .map {case (k, v) => ((k, v).asInstanceOf[T], rightPairs.get(k).get)}.toIterator
+  }
+  
+  private def computeWithFunc(split: Partition, context: TaskContext, region: Region[K, V]): Iterator[(T, V)] = {
+    val leftPairs = left.iterator(split, context).toList.map(t => (t, func(t)))
+    val leftKeys = leftPairs.map { case (t, k) => k}.toSet
+    // Note: get all will return (key, null) for non-exist entry, so remove those entries
+    val rightPairs = region.getAll(leftKeys).filter { case (k, v) => v != null}
+    leftPairs.filter { case (t, k) => rightPairs.contains(k)}.map {case (t, k) => (t, rightPairs.get(k).get)}.toIterator
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireOuterJoinRDD.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireOuterJoinRDD.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireOuterJoinRDD.scala
new file mode 100644
index 0000000..ec3a512
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireOuterJoinRDD.scala
@@ -0,0 +1,69 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import com.gemstone.gemfire.cache.Region
+import io.pivotal.gemfire.spark.connector.GemFireConnectionConf
+import org.apache.spark.{TaskContext, Partition}
+import org.apache.spark.rdd.RDD
+import scala.collection.JavaConversions._
+
+/**
+ * An `RDD[ T, Option[V] ]` that represents the result of a left outer join 
+ * between `left` RDD[T] and the specified GemFire Region[K, V].
+ */
+class GemFireOuterJoinRDD[T, K, V] private[connector]
+ ( left: RDD[T],
+   func: T => K,
+   val regionPath: String,
+   val connConf: GemFireConnectionConf
+  ) extends RDD[(T, Option[V])](left.context, left.dependencies) {
+
+  /** validate region existence when GemFireRDD object is created */
+  validate()
+
+  /** Validate region, and make sure it exists. */
+  private def validate(): Unit = connConf.getConnection.validateRegion[K, V](regionPath)
+
+  override protected def getPartitions: Array[Partition] = left.partitions
+
+  override def compute(split: Partition, context: TaskContext): Iterator[(T, Option[V])] = {
+    val region = connConf.getConnection.getRegionProxy[K, V](regionPath)
+    if (func == null) computeWithoutFunc(split, context, region)
+    else computeWithFunc(split, context, region)
+  }
+
+  /** T is (K1, V1), and K1 and K are the same type since there's no map function `func` */
+  private def computeWithoutFunc(split: Partition, context: TaskContext, region: Region[K, V]): Iterator[(T, Option[V])] = {
+    val leftPairs = left.iterator(split, context).toList.asInstanceOf[List[(K, _)]]
+    val leftKeys = leftPairs.map { case (k, v) => k}.toSet
+    // Note: get all will return (key, null) for non-exist entry
+    val rightPairs = region.getAll(leftKeys)
+    // rightPairs is a java.util.Map, not scala map, so need to convert map.get() to Option
+    leftPairs.map{ case (k, v) => ((k, v).asInstanceOf[T], Option(rightPairs.get(k))) }.toIterator
+  }
+
+  private def computeWithFunc(split: Partition, context: TaskContext, region: Region[K, V]): Iterator[(T, Option[V])] = {
+    val leftPairs = left.iterator(split, context).toList.map(t => (t, func(t)))
+    val leftKeys = leftPairs.map { case (t, k) => k}.toSet
+    // Note: get all will return (key, null) for non-exist entry
+    val rightPairs = region.getAll(leftKeys)
+    // rightPairs is a java.util.Map, not scala map, so need to convert map.get() to Option
+    leftPairs.map{ case (t, k) => (t, Option(rightPairs.get(k)))}.toIterator
+  }
+}
+

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartition.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartition.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartition.scala
new file mode 100644
index 0000000..02e1eed
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartition.scala
@@ -0,0 +1,36 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import org.apache.spark.Partition
+
+/**
+ * This serializable class represents a GemFireRDD partition. Each partition is mapped
+ * to one or more buckets of region. The GemFireRDD can materialize the data of the 
+ * partition based on all information contained here.
+ * @param partitionId partition id, a 0 based number.
+ * @param bucketSet region bucket id set for this partition. Set.empty means whole
+ *                  region (used for replicated region)
+ * @param locations preferred location for this partition                  
+ */
+case class GemFireRDDPartition (
+  partitionId: Int, bucketSet: Set[Int], locations: Seq[String] = Nil)
+  extends Partition  {
+  
+  override def index: Int = partitionId
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitioner.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitioner.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitioner.scala
new file mode 100644
index 0000000..807a979
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitioner.scala
@@ -0,0 +1,59 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import io.pivotal.gemfire.spark.connector.GemFireConnection
+import io.pivotal.gemfire.spark.connector.internal.RegionMetadata
+import org.apache.spark.{Logging, Partition}
+
+import scala.reflect.ClassTag
+
+/**
+ * A GemFireRDD partitioner is used to partition the region into multiple RDD partitions.
+ */
+trait GemFireRDDPartitioner extends Serializable {
+
+  def name: String
+  
+  /** the function that generates partitions */
+  def partitions[K: ClassTag, V: ClassTag]
+    (conn: GemFireConnection, md: RegionMetadata, env: Map[String, String]): Array[Partition]
+}
+
+object GemFireRDDPartitioner extends Logging {
+
+  /** To add new partitioner, just add it to the following list */
+  final val partitioners: Map[String, GemFireRDDPartitioner] =
+    List(OnePartitionPartitioner, ServerSplitsPartitioner).map(e => (e.name, e)).toMap
+
+  /**
+   * Get a partitioner based on given name, a default partitioner will be returned if there's
+   * no partitioner for the given name. 
+   */
+  def apply(name: String = defaultPartitionedRegionPartitioner.name): GemFireRDDPartitioner = {
+    val p = partitioners.get(name)
+    if (p.isDefined) p.get else {
+      logWarning(s"Invalid preferred partitioner name $name.")
+      defaultPartitionedRegionPartitioner
+    }
+  }
+
+  val defaultReplicatedRegionPartitioner = OnePartitionPartitioner
+
+  val defaultPartitionedRegionPartitioner = ServerSplitsPartitioner
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitionerImpl.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitionerImpl.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitionerImpl.scala
new file mode 100644
index 0000000..72904ee
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDPartitionerImpl.scala
@@ -0,0 +1,89 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import io.pivotal.gemfire.spark.connector.GemFireConnection
+import io.pivotal.gemfire.spark.connector.internal.RegionMetadata
+import io.pivotal.gemfire.spark.connector.NumberPartitionsPerServerPropKey
+import org.apache.spark.Partition
+import scala.collection.JavaConversions._
+import scala.collection.immutable.SortedSet
+import scala.collection.mutable
+import scala.reflect.ClassTag
+
+/** This partitioner maps whole region to one GemFireRDDPartition */
+object OnePartitionPartitioner extends GemFireRDDPartitioner {
+
+  override val name = "OnePartition"
+
+  override def partitions[K: ClassTag, V: ClassTag]
+    (conn: GemFireConnection, md: RegionMetadata, env: Map[String, String]): Array[Partition] =
+    Array[Partition](new GemFireRDDPartition(0, Set.empty))
+}
+
+/**
+  * This partitioner maps whole region to N * M GemFire RDD partitions, where M is the number of 
+  * GemFire servers that contain the data for the given region. Th default value of N is 1.
+  */
+object ServerSplitsPartitioner extends GemFireRDDPartitioner {
+
+  override val name = "ServerSplits"
+
+  override def partitions[K: ClassTag, V: ClassTag]
+  (conn: GemFireConnection, md: RegionMetadata, env: Map[String, String]): Array[Partition] = {
+    if (md == null) throw new RuntimeException("RegionMetadata is null")
+    val n = try { env.getOrElse(NumberPartitionsPerServerPropKey, "2").toInt } catch { case e: NumberFormatException => 2 }
+    if (!md.isPartitioned || md.getServerBucketMap == null || md.getServerBucketMap.isEmpty)
+      Array[Partition](new GemFireRDDPartition(0, Set.empty))
+    else {
+      val map = mapAsScalaMap(md.getServerBucketMap)
+        .map { case (srv, set) => (srv, asScalaSet(set).map(_.toInt)) }.toList
+        .map { case (srv, set) => (srv.getHostName, set) }
+       doPartitions(map, md.getTotalBuckets, n)
+    }
+  }
+
+  /** Converts server to bucket ID set list to array of RDD partitions */
+  def doPartitions(serverBucketMap: List[(String, mutable.Set[Int])], totalBuckets: Int, n: Int)
+    : Array[Partition] = {
+
+    // method that calculates the group size for splitting "k" items into "g" groups
+    def groupSize(k: Int, g: Int): Int = scala.math.ceil(k / g.toDouble).toInt
+
+    // 1. convert list of server and bucket set pairs to a list of server and sorted bucket set pairs
+    val srvToSortedBucketSet = serverBucketMap.map { case (srv, set) => (srv, SortedSet[Int]() ++ set) }
+
+    // 2. split bucket set of each server into n splits if possible, and server to Seq(server)
+    val srvToSplitedBuckeSet = srvToSortedBucketSet.flatMap { case (host, set) =>
+      if (set.isEmpty) Nil else set.grouped(groupSize(set.size, n)).toList.map(s => (Seq(host), s)) }
+
+    // 3. calculate empty bucket IDs by removing all bucket sets of all servers from the full bucket sets
+    val emptyIDs = SortedSet[Int]() ++ ((0 until totalBuckets).toSet /: srvToSortedBucketSet) {case (s1, (k, s2)) => s1 &~ s2}
+
+    // 4. distribute empty bucket IDs to all partitions evenly.
+    //    The empty buckets do not contain data when partitions are created, but they may contain data
+    //    when RDD is materialized, so need to include those bucket IDs in the partitions.
+    val srvToFinalBucketSet = if (emptyIDs.isEmpty) srvToSplitedBuckeSet
+      else srvToSplitedBuckeSet.zipAll(
+        emptyIDs.grouped(groupSize(emptyIDs.size, srvToSplitedBuckeSet.size)).toList, (Nil, Set.empty), Set.empty).map
+          { case ((server, set1), set2) => (server, SortedSet[Int]() ++ set1 ++ set2) }
+
+    // 5. create array of partitions w/ 0-based index
+    (0 until srvToFinalBucketSet.size).toList.zip(srvToFinalBucketSet).map
+      { case (i, (srv, set)) => new GemFireRDDPartition(i, set, srv) }.toArray
+  }
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDWriter.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDWriter.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDWriter.scala
new file mode 100644
index 0000000..b6b1330
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRDDWriter.scala
@@ -0,0 +1,82 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import com.gemstone.gemfire.cache.Region
+import io.pivotal.gemfire.spark.connector._
+import org.apache.spark.{Logging, TaskContext}
+
+import scala.collection.Iterator
+import java.util.{HashMap => JMap}
+
+/** This trait provide some common code for pair and non-pair RDD writer */
+private[rdd] abstract class GemFireRDDWriterBase (opConf: Map[String, String]) extends Serializable {
+
+  val batchSize = try { opConf.getOrElse(RDDSaveBatchSizePropKey, RDDSaveBatchSizeDefault.toString).toInt}
+                  catch { case e: NumberFormatException => RDDSaveBatchSizeDefault }
+
+  def mapDump(map: Map[_, _], num: Int): String = {
+    val firstNum = map.take(num + 1)
+    if (firstNum.size > num) s"$firstNum ..." else s"$firstNum"    
+  }  
+}
+
+/**
+ * Writer object that provides write function that saves non-pair RDD partitions to GemFire.
+ * Those functions will be executed on Spark executors.
+ * @param regionPath the full path of the region where the data is written to
+ */
+class GemFireRDDWriter[T, K, V] 
+  (regionPath: String, connConf: GemFireConnectionConf, opConf: Map[String, String] = Map.empty)
+  extends GemFireRDDWriterBase(opConf) with Serializable with Logging {
+
+  def write(func: T => (K, V))(taskContext: TaskContext, data: Iterator[T]): Unit = {
+    val region: Region[K, V] = connConf.getConnection.getRegionProxy[K, V](regionPath)
+    var count = 0
+    val chunks = data.grouped(batchSize)
+    chunks.foreach { chunk =>
+      val map = chunk.foldLeft(new JMap[K, V]()){case (m, t) => val (k, v) = func(t); m.put(k, v); m}
+      region.putAll(map)
+      count += chunk.length
+    }
+    logDebug(s"$count entries (batch.size = $batchSize) are saved to region $regionPath")
+  }
+}
+
+
+/**
+ * Writer object that provides write function that saves pair RDD partitions to GemFire.
+ * Those functions will be executed on Spark executors.
+ * @param regionPath the full path of the region where the data is written to
+ */
+class GemFirePairRDDWriter[K, V]
+  (regionPath: String, connConf: GemFireConnectionConf, opConf: Map[String, String] = Map.empty)
+  extends GemFireRDDWriterBase(opConf) with Serializable with Logging {
+
+  def write(taskContext: TaskContext, data: Iterator[(K, V)]): Unit = {
+    val region: Region[K, V] = connConf.getConnection.getRegionProxy[K, V](regionPath)
+    var count = 0
+    val chunks = data.grouped(batchSize)
+    chunks.foreach { chunk =>
+      val map = chunk.foldLeft(new JMap[K, V]()){case (m, (k,v)) => m.put(k,v); m}
+      region.putAll(map)
+      count += chunk.length
+    }
+    logDebug(s"$count entries (batch.batch = $batchSize) are saved to region $regionPath")
+  }
+}
+

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRegionRDD.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRegionRDD.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRegionRDD.scala
new file mode 100644
index 0000000..4535917
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/internal/rdd/GemFireRegionRDD.scala
@@ -0,0 +1,138 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.internal.rdd
+
+import scala.collection.Seq
+import scala.reflect.ClassTag
+import org.apache.spark.rdd.RDD
+import org.apache.spark.{TaskContext, Partition, SparkContext}
+import io.pivotal.gemfire.spark.connector.{GemFireConnectionConf, PreferredPartitionerPropKey}
+import io.pivotal.gemfire.spark.connector.internal.rdd.GemFireRDDPartitioner._
+
+/**
+ * This class exposes GemFire region as a RDD.
+ * @param sc the Spark Context
+ * @param regionPath the full path of the region
+ * @param connConf the GemFireConnectionConf to access the region
+ * @param opConf the parameters for this operation, such as preferred partitioner.
+ */
+class GemFireRegionRDD[K, V] private[connector]
+  (@transient sc: SparkContext,
+   val regionPath: String,
+   val connConf: GemFireConnectionConf,
+   val opConf: Map[String, String] = Map.empty,
+   val whereClause: Option[String] = None 
+  ) (implicit ctk: ClassTag[K], ctv: ClassTag[V])
+  extends RDD[(K, V)](sc, Seq.empty) {
+
+  /** validate region existence when GemFireRDD object is created */
+  validate()
+
+  /** Validate region, and make sure it exists. */
+  private def validate(): Unit = connConf.getConnection.validateRegion[K, V](regionPath)
+
+  def kClassTag = ctk
+  
+  def vClassTag = ctv
+
+  /**
+   * method `copy` is used by method `where` that creates new immutable
+   * GemFireRDD instance based this instance.
+   */
+  private def copy(
+    regionPath: String = regionPath,
+    connConf: GemFireConnectionConf = connConf,
+    opConf: Map[String, String] = opConf,
+    whereClause: Option[String] = None
+  ): GemFireRegionRDD[K, V] = {
+
+    require(sc != null,
+    """RDD transformation requires a non-null SparkContext. Unfortunately
+      |SparkContext in this GemFireRDD is null. This can happen after 
+      |GemFireRDD has been deserialized. SparkContext is not Serializable,
+      |therefore it deserializes to null. RDD transformations are not allowed
+      |inside lambdas used in other RDD transformations.""".stripMargin )
+
+    new GemFireRegionRDD[K, V](sc, regionPath, connConf, opConf, whereClause)
+  }
+
+  /** When where clause is specified, OQL query
+    * `select key, value from /<region-path>.entries where <where clause> `
+    * is used to filter the dataset.
+    */
+  def where(whereClause: Option[String]): GemFireRegionRDD[K, V] = {
+    if (whereClause.isDefined) copy(whereClause = whereClause)
+    else this
+  }
+
+  /** this version is for Java API that doesn't use scala.Option */
+  def where(whereClause: String): GemFireRegionRDD[K, V] = {
+    if (whereClause == null || whereClause.trim.isEmpty) this
+    else copy(whereClause = Option(whereClause.trim))
+  }
+
+  /**
+   * Use preferred partitioner generate partitions. `defaultReplicatedRegionPartitioner`
+   * will be used if it's a replicated region. 
+   */
+  override def getPartitions: Array[Partition] = {
+    val conn = connConf.getConnection
+    val md = conn.getRegionMetadata[K, V](regionPath)
+    md match {
+      case None => throw new RuntimeException(s"region $regionPath was not found.")
+      case Some(data) =>
+        logInfo(s"""RDD id=${this.id} region=$regionPath conn=${connConf.locators.mkString(",")}, env=$opConf""")
+        val p = if (data.isPartitioned) preferredPartitioner else defaultReplicatedRegionPartitioner
+        val splits = p.partitions[K, V](conn, data, opConf)
+        logDebug(s"""RDD id=${this.id} region=$regionPath partitions=\n  ${splits.mkString("\n  ")}""")
+        splits
+    }
+  }
+
+  /**
+   * provide preferred location(s) (host name(s)) of the given partition. 
+   * Only some partitioner implementation(s) provides this info, which is
+   * useful when Spark cluster and GemFire cluster share some hosts.
+   */
+  override def getPreferredLocations(split: Partition) =
+    split.asInstanceOf[GemFireRDDPartition].locations
+
+  /**
+   * Get preferred partitioner. return `defaultPartitionedRegionPartitioner` if none
+   * preference is specified. 
+   */
+  private def preferredPartitioner = 
+    GemFireRDDPartitioner(opConf.getOrElse(
+      PreferredPartitionerPropKey, GemFireRDDPartitioner.defaultPartitionedRegionPartitioner.name))
+
+  /** materialize a RDD partition */
+  override def compute(split: Partition, context: TaskContext): Iterator[(K, V)] = {
+    val partition = split.asInstanceOf[GemFireRDDPartition]
+    logDebug(s"compute RDD id=${this.id} partition $partition")
+    connConf.getConnection.getRegionData[K,V](regionPath, whereClause, partition)
+    // new InterruptibleIterator(context, split.asInstanceOf[GemFireRDDPartition[K, V]].iterator)
+  }
+}
+
+object GemFireRegionRDD {
+
+  def apply[K: ClassTag, V: ClassTag](sc: SparkContext, regionPath: String,
+    connConf: GemFireConnectionConf, opConf: Map[String, String] = Map.empty)
+    : GemFireRegionRDD[K, V] =
+    new GemFireRegionRDD[K, V](sc, regionPath, connConf, opConf)
+
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/GemFireJavaRegionRDD.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/GemFireJavaRegionRDD.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/GemFireJavaRegionRDD.scala
new file mode 100644
index 0000000..1a2e8f2
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/GemFireJavaRegionRDD.scala
@@ -0,0 +1,26 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.javaapi
+
+import io.pivotal.gemfire.spark.connector.internal.rdd.GemFireRegionRDD
+import org.apache.spark.api.java.JavaPairRDD
+
+class GemFireJavaRegionRDD[K, V](rdd: GemFireRegionRDD[K, V]) extends JavaPairRDD[K, V](rdd)(rdd.kClassTag, rdd.vClassTag) {
+  
+  def where(whereClause: String): GemFireJavaRegionRDD[K, V] = new GemFireJavaRegionRDD(rdd.where(whereClause))
+  
+}

http://git-wip-us.apache.org/repos/asf/incubator-geode/blob/ddee87fe/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/JavaAPIHelper.scala
----------------------------------------------------------------------
diff --git a/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/JavaAPIHelper.scala b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/JavaAPIHelper.scala
new file mode 100644
index 0000000..3d10fb8
--- /dev/null
+++ b/geode-spark-connector/gemfire-spark-connector/src/main/scala/io/pivotal/gemfire/spark/connector/javaapi/JavaAPIHelper.scala
@@ -0,0 +1,53 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package io.pivotal.gemfire.spark.connector.javaapi
+
+import org.apache.spark.api.java.{JavaPairRDD, JavaRDD}
+import org.apache.spark.streaming.api.java.{JavaPairDStream, JavaDStream}
+
+import scala.reflect.ClassTag
+import scala.collection.JavaConversions._
+
+/**
+ *  A helper class to make it possible to access components written in Scala from Java code.
+ */
+private[connector] object JavaAPIHelper {
+
+  /** Returns a `ClassTag` of a given runtime class. */
+  def getClassTag[T](clazz: Class[T]): ClassTag[T] = ClassTag(clazz)
+
+  /**
+   * Produces a ClassTag[T], which is actually just a casted ClassTag[AnyRef].
+   * see JavaSparkContext.fakeClassTag in Spark for more info.
+   */
+  def fakeClassTag[T]: ClassTag[T] = ClassTag.AnyRef.asInstanceOf[ClassTag[T]]
+
+  /** Converts a Java `Properties` to a Scala immutable `Map[String, String]`. */
+  def propertiesToScalaMap[K, V](props: java.util.Properties): Map[String, String] =
+    Map(props.toSeq: _*)
+
+  /** convert a JavaRDD[(K,V)] to JavaPairRDD[K,V] */
+  def toJavaPairRDD[K, V](rdd: JavaRDD[(K, V)]): JavaPairRDD[K, V] =
+    JavaPairRDD.fromJavaRDD(rdd)
+
+  /** convert a JavaDStream[(K,V)] to JavaPairDStream[K,V] */
+  def toJavaPairDStream[K, V](ds: JavaDStream[(K, V)]): JavaPairDStream[K, V] =
+    JavaPairDStream.fromJavaDStream(ds)
+
+  /** an empty Map[String, String] for default opConf **/
+  val emptyStrStrMap: Map[String, String] = Map.empty
+}


Mime
View raw message