如何在不中断的情况下运行火花流

Question

i'm trying to save save tweets from twitter with help twitter-streaming.But i have one problem: my program stops working after some period of time(depends on Batch Interval for 1 millis near 4-5 sec). So, could you help me with this problem solving). Tell me please what is wrong?

When batch interval near 100 millis i see some records like

19/08/06 23:45:26 INFO BlockRDD: Removing RDD 103 from persistence list
19/08/06 23:45:26 INFO BlockManager: Removing RDD 103
19/08/06 23:45:26 INFO TwitterInputDStream: Removing blocks of RDD BlockRDD[103] at createStream at Twitter.java:35 of time 1565124324340 ms
19/08/06 23:45:26 INFO ReceivedBlockTracker: Deleting batches: 1565124324320 ms
19/08/06 23:45:26 INFO InputInfoTracker: remove old batch metadata: 1565124324320 ms
-------------------------------------------
Time: 1565124325500 ms

When batch interval is "big" and any data isn't avaible, i just see message abou Spark UI starting and finished.

package TwitterAnalysis;

import org.apache.spark.*;
import org.apache.spark.storage.StorageLevel;
import org.apache.spark.streaming.*;
import org.apache.spark.streaming.api.java.*;
import org.apache.spark.streaming.twitter.*;

import twitter4j.Status;



public class Twitter {

    private static void setTwitterOAuth() {
        System.setProperty("twitter4j.oauth.consumerKey", TwitterOAuthKey.consumerKey);
        System.setProperty("twitter4j.oauth.consumerSecret", TwitterOAuthKey.consumerSecret);
        System.setProperty("twitter4j.oauth.accessToken", TwitterOAuthKey.accessToken);
        System.setProperty("twitter4j.oauth.accessTokenSecret", TwitterOAuthKey.accessTokenSecret);
    }



    public static void main(String [] args) {

        setTwitterOAuth();

        SparkConf conf = new SparkConf().setMaster("local[*]")
                                         .setAppName("SparkTwitter");

      //  JavaSparkContext sparkContext = new JavaSparkContext(conf);
        JavaStreamingContext jssc = new JavaStreamingContext(conf, new Duration(10000));


        JavaReceiverInputDStream<Status> twitterStream = TwitterUtils.createStream(jssc);

        //Stream that contains just tweets in english
        JavaDStream<Status> enTweetsDStream=twitterStream.filter((status) -> "en".equalsIgnoreCase(status.getLang()));
        enTweetsDStream.persist(StorageLevel.MEMORY_AND_DISK());


        enTweetsDStream.print();
        jssc.start();


    }

}

Answer 1

根据此答案： Spark 2.0.0的Twitter流驱动程序不再可用，在Spark 2.0及更高版本上没有可用的twitter-streaming-driver。解决方案请选择Spark的早期版本）