Skip navigation


Filter by Answers and Ideas
Hi, A spark application is receiving data from hundreds of topics. For a particular RDD received from a batch of 5 second,  i want to groupby messages based on topic name and put them in different mardb table based on topic name. Groupby method is consuming time and advice not to be used. Please suggest any other method if there any.

Hi,  I have created a sample spark application which reads the data from MapR stream and save it to MapR DB. I get below error while saving data to MapR DB. I am using Spark Version 2.1.0 and MapR DB version as 5.2.2.   Exception in thread "streaming-job-executor-0" java.lang.NoClassDefFoundError: com/mapr/db/impl/MapRDBImpl
Top & Trending
I am trying to use Spark Streaming with MapR Streams. Streaming application is enable checkoonting. I can execute Streaming application,but not restore from checkpoint with following exception. org.apache.kafka.common.errors.UnknownTopicOrPartitionException :No such file or directory (2) Could not seek.   And I found the following
Filter by Training Content
Hello All,  currently, I am preparing for MapR Hadoop certification exam and as a part of that I came across the Lesson 2 Lab where i am supposed to run a yarn command to launch a YARN job but i could not get desired result from that command. Please help me if I am missing any thing.    The command is yarn jar
Top & Trending
Hi, We are using a 2 node development cluster for about 3 months and MCS gives me this alarm today (Installation Directory Full Alarm).     Upon checking the /opt directory I noticed that fluentd consume a lot of space only for the logs itself.   Is it ok to delete this one ? or for future references can I just move it to