port 8080 already in use

[WangYongNingDA]

i have deploy griffin and run at local,but when i run java -jar service/target/service.jar, it says port 8080 is used.how can i change the port to others?

[bhlx3lyx7]

The service is based on spring boot, you can change the port like other normal spring boot applications.

Sent from Mail for Windows 10

From: WangYongNingDA Sent: 2017年12月13日 14:02 To: eBay/griffin Cc: Subscribed Subject: [eBay/griffin] port 8080 already in use (#50)

i have deploy griffin and run at local,but when i run java -jar service/target/service.jar, it says port 8080 is used.how can i change the port to others? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[WangYongNingDA]

hi bhlx3lyx7 thank you,I have reslove this problem.but when i use livy to submit sparkjob,the errors are as below:

Warning: Skip remote jar hdfs://wecloud-cluster/project/pgxl/griffin/griffin-measure.jar.
Warning: Skip remote jar hdfs://wecloud-cluster/project/pgxl/griffin/datanucleus-api-jdo-3.2.6.jar.
Warning: Skip remote jar hdfs://wecloud-cluster/project/pgxl/griffin/datanucleus-core-3.2.10.jar.
Warning: Skip remote jar hdfs://wecloud-cluster/project/pgxl/griffin/datanucleus-rdbms-3.2.9.jar.
java.lang.ClassNotFoundException: org.apache.griffin.measure.Application
        at java.net.URLClassLoader$1.run(URLClassLoader.java:366)
        at java.net.URLClassLoader$1.run(URLClassLoader.java:355)
        at java.security.AccessController.doPrivileged(Native Method)
        at java.net.URLClassLoader.findClass(URLClassLoader.java:354)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:425)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:358)
        at java.lang.Class.forName0(Native Method)
        at java.lang.Class.forName(Class.java:270)
        at org.apache.spark.util.Utils$.classForName(Utils.scala:175)
        at org.apache.spark.deploy.SparkSubmit$.org$apache$spark$deploy$SparkSubmit$$runMain(SparkSubmit.scala:689)
        at org.apache.spark.deploy.SparkSubmit$.doRunMain$1(SparkSubmit.scala:181)
        at org.apache.spark.deploy.SparkSubmit$.submit(SparkSubmit.scala:206)
        at org.apache.spark.deploy.SparkSubmit$.main(SparkSubmit.scala:121)
        at org.apache.spark.deploy.SparkSubmit.main(SparkSubmit.scala)

[bhlx3lyx7]

It seems like to be some version issue, the griffin-measure.jar is using scala 2.10 by default, which means the spark version should be matched with scala 2.10. Griffin measure works on spark 1.6, but the scala version could be modified, you can package your own jar for scala 2.11, just modify scala.version and scala.binary.version in the pom.xml of measure module, run mvn clean package in the measure directory.

[WangYongNingDA]

Thanks for your response,i have resloved the problem by change the config.but now the job can be submitted to the yarn cluser,other problems comes. According to the confs in the sparkJob.properties,sparkJob.spark.jars.packages is "com.databricks:spark-avro_2.10:2.0.1". When I submit jobs by using "--packages com.databricks:spark-avro_2.10:2.0.1 ",it works well.But when use livy to submit jobs,the errors comes:

Application application_1512628890181_92576 failed 2 times due to AM Container for appattempt_1512628890181_92576_000002 exited with exitCode: -1000
For more detailed output, check application tracking page:http://xy180-wecloud-198:8088/proxy/application_1512628890181_92576/Then, click on links to logs of each attempt.
Diagnostics: File does not exist: hdfs://wecloud-cluster/user/pgxl/.sparkStaging/application_1512628890181_92576/com.databricks_spark-avro_2.10-2.0.1.jar
java.io.FileNotFoundException: File does not exist: hdfs://wecloud-cluster/user/pgxl/.sparkStaging/application_1512628890181_92576/com.databricks_spark-avro_2.10-2.0.1.jar

I have delete the code about spark.jars.packages in the class SparkSubmitJob,it also don't works.

[bhlx3lyx7]

It's a tricky issue, the measure module supports avro file by this package, but we don't want to package it in measure.jar, so we need to tell spark where it is. sparkJob.spark.jars.packages=com.databricks:spark-avro_2.10:2.0.1 will let spark application download this package when it starts, it exits if download fails. You can download this package manually, put it in hdfs, and set the config in sparkJob.properties, like sparkJob.jars_4=hdfs:///path/to/spark-avro_2.10.jar, then modify some code in service module to enable this property when submit to livy, that will resolve this issue. In the next version, we'll consider to modify sparkJob.properties, to fit livy parameters friendly.