About mjohansson

mjohansson · ‎09-14-2016

Sorry, too quick on the trigger. Normally SPARK_HOME=/usr/hdp/current/spark-client In my (old) sandbox the start_master.sh script is located at /usr/hdp/2.4.0.0-169/spark/sbin/start-master.sh.

mjohansson · ‎09-14-2016

Normally it's under /usr/hdp/current/spark-client

mjohansson · ‎06-14-2016

Hi, Not sure what you mean with Spark on R? If that is RHadoop, MS R Server or DistributedR those are not really comparable just like that with SparkR, ie no drop-in replacement, as the functionality differs. So before doing a performance comparison you have to look into which functionality you would like to compare. /Best regards, Mats

mjohansson · ‎06-14-2016

Sounds ok as long as you continue to use /usr/hdp as the directory path for the new drive and shut down the cluster/node while doing the change.

mjohansson · ‎06-14-2016

Hi @Uday Vakalapudi, HDP do not require any special sizing for /dev as long as we follow the OS recommendations. The "baseline" we present includes general Linux requirements for system partitions. So in this case I recommend to follow the requirements for /dev issued by the operating system vendor.

mjohansson · ‎05-10-2016

@Predrag Minovic - Is it possible to remove the auditlog in hdfs as well without affecting lineage etc?

mjohansson · ‎05-09-2016

Is there any best practices on how to handle the ranger_audit.xa_access_audit table over time when it's growing big. In particular can we truncate/delete rows without affecting lineage?

mjohansson · ‎04-25-2016

Looks like hadoop-yarn-common-x.x.x.jar is missing.

mjohansson · ‎04-08-2016

Hi @Roberto Sancho Please have a look at this post? https://community.hortonworks.com/questions/9762/how-to-get-rid-of-stale-alerts-in-ambari.html

mjohansson · ‎04-08-2016

Hi, http://<ambari-server>:8080/api/v1/stacks/<your stack>/versions/<your-version>/services/FALCON For example: http://sb133:8080/api/v1/stacks/HDP/versions/2.4/services/FALCON

Online	Offline
Last Visited	‎08-14-2019 09:54 AM

Member Since	‎09-25-2015 07:44 AM
Last Visited	‎08-14-2019 09:54 AM
Posts	72
Kudos received	60

Cloudera Community

Re: HDFS_PATH in Apache Atlas Search

Re: How does the input port know which device data...

Re: HIVE/HBASE/SOLR

Re: Hive database replication with MySQL Cluster

Re: Can there be any incompatibility issues with E...

Re: Where is $SPARK_HOME ?

Re: Where is $SPARK_HOME ?

Re: Spark on R vs R on Spark (SparkR) ?

Re: Moving /usr/hdp to a different drive

Re: Why do we allocate 32 GB (recommended) for /d...

Re: Can the ranger_audit.xa_access_audit table be ...

Can the ranger_audit.xa_access_audit table be trun...

Re: JARS are missing, can you please let me know w...

Re: ambari alerts-

Re: REST API to get individual HDP component versi...