My Oracle Support Banner

During Mammoth Migration on the Host with Cloudera Manager Role, StartHadoopServices Fails Adding the Spark2 Service at Addspark2/Exec[add_spark2] (Doc ID 2861131.1)

Last updated on APRIL 12, 2022

Applies to:

Big Data Appliance Integrated Software - Version 4.13.0 and later
x86_64

Symptoms

NOTE: In the examples that follow, user details, cluster names, hostnames, directory paths, filenames, etc. represent a fictitious sample (and are used to provide an illustrative example only). Any similarity to actual persons, or entities, living or dead, is purely coincidental and not intended in any manner.

Performing a mammoth migration for the host with Cloudera Manager (CM) role, Node 3 by default, on a BDA cluster with AD Kerberos following:
Node 3 Migration and Recommission on Oracle Big Data Appliance V4.11 and Higher (Doc ID 2524859.1)
or
Node 3 Migration and Reprovision on Oracle Big Data Appliance V4.1 OL6 Hadoop Cluster to Manage a Hardware Failure (Doc ID 1984854.1)

fails in Step 4, StartHadoopServices, adding the Spark 2 service with errors like below:

************************************
Error [16892]: (//<HOSTNAME_TARGET_CMHOST>.<DOMAIN>//Stage[addsvc]/Hadoop::Addspark2/Exec[add_spark2]/returns) change from notrun to 0 failed:
/opt/oracle/BDAMammoth/bdaconfig/tmp/addspark2.sh &>
/opt/oracle/BDAMammoth/bdaconfig/tmp/addspark2_1649101431.out returned 1 instead of one of [0]
************************************

In the case of a BDA cluster which migrated from OL6 to OL7, confirm that the OL7 version of the Spark 2 parcel is distributed on the target host i.e. the host where the original Node 3 is migrating to.

Check Hosts > Parcels in CM to see if there is an error distributing the OL7 version of the Spark 2 parcel.

Errors can look like below based on BDA 4.13. The Spark 2 parcel version can vary with BDA release.

Error while downloading parcel SPARK2-2.3.0.cloudera3-1.cdh5.13.3.p0.458809-el7.parcel: Parcel SPARK2-2.3.0.cloudera3-1.cdh5.13.3.p0.458809-el7.parcel not found on remote repository.

Note: Ignore any errors about the parcel not distributing to the original host with CM role.  This host is down because it is being migrated to a new i.e. target host. Such errors are like:

Error for parcel SPARK2-2.3.0.cloudera3-1.cdh5.13.3.p0.458809-unknown : One or more hosts did not report their OS Distribution. This is likely due torunning an outdated version of the Cloudera Manager Agent. Please run the Host Inspector to check the Agent versions.

Error when distributing to <HOSTNAME_TARGET_CMHOST>.<DOMAIN>: Host is in bad health.

 

 

Cause

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


In this Document
Symptoms
Cause
Solution
 Install Spark 2 Parcel on Target Host
 Setup the Spark 2 CSD file
 Install Spark 2
 Resume migration
References


My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.