我必须映射一个表,其中写入了一个应用程序的使用历史。这个表有这些元组:
<AppId,date,cpuUsage,memoryUsage>
<AppId,date,cpuUsage,memoryUsage>
<AppId,date,cpuUsage,memoryUsage>
<AppId,date,cpuUsage,memoryUsage>
<AppId,date,cpuUsage,memoryUsage>
AppId
总是不同的,因为在许多应用程序中被引用,date
用这种格式表示,dd/mm/yyyy hh/mm
、cpuUsage
和memoryUsage
用%
表示,所以例如:
<3ghffh3t482age20304,230720142245,0.2,3,5>
我以这种方式从cassandra检索数据(小片段):
public static void main(String[] args) {
Cluster cluster;
Session session;
cluster = Cluster.builder().addContactPoint("127.0.0.1").build();
session = cluster.connect();
session.execute("CREATE KEYSPACE IF NOT EXISTS foo WITH replication "
+ "= {'class':'SimpleStrategy', 'replication_factor':3};");
String createTableAppUsage = "CREATE TABLE IF NOT EXISTS foo.appusage"
+ "(appid text,date text, cpuusage double, memoryusage double, "
+ "PRIMARY KEY(appid,date) " + "WITH CLUSTERING ORDER BY (time ASC);";
session.execute(createTableAppUsage);
// Use select to get the appusage's table rows
ResultSet resultForAppUsage = session.execute("SELECT appid,cpuusage FROM foo.appusage");
for (Row row: resultForAppUsage)
System.out.println("appid :" + row.getString("appid") +" "+ "cpuusage"+row.getString("cpuusage"));
// Clean up the connection by closing it
cluster.close();
}
所以,我现在的问题是通过key value
映射数据并创建一个集成此代码的元组(不工作的代码段):
<AppId,cpuusage>
JavaPairRDD<String, Integer> saveTupleKeyValue =someStructureFromTakeData.mapToPair(new PairFunction<String, String, Integer>() {
public Tuple2<String, Integer> call(String x) {
return new Tuple2(x, y);
}
如何使用RDD和reduce eg. cpuusage >50
映射appId和cpuusage ?
帮忙吗?
假设您已经创建了一个有效的SparkContext sparkContext
,并将spark-cassandra连接器依赖项添加到您的项目中,并配置了spark应用程序以与cassandra集群通信(请参阅文档),那么我们可以像这样在RDD中加载数据:
val data = sparkContext.cassandraTable("foo", "appusage").select("appid", "cpuusage")
在Java中,这个想法是相同的,但它需要更多的管道,这里描述