Question-106: Can you have different version of the JDK across the host in CDH cluster?

Answer:  No, you can use different JDK like Oracle provided JDK, or open JDK.  But same version of JDK must be installed on each cluster host.  As mentioned previously, it is highly recommended that you install the JDK at /usr/java/jdk-version

Question-107:  What happens when you use G1GC garbage collector of Java?

Answer:  When using G1GC garbage collection, then pauses for garbage collection are shorter, component will usually be more responsive, but they are more sensitive to JVMs with over committed memory usage. 

Question-108:  While installing Cloudera manager server, why it is recommended to enable AUTO-TLS?

Answer: Because auto TLS simplify the process of enabling and managing TLS encryption on your cluster. It automates the creation of an internal certificate authority and deployment of certificate across all cluster host. It can also automate the distribution of existing certificate, such as those signed by a public certificate authority.  So whenever you add a new host or service  to your cluster with auto TLS enabled,  it automatically create and deploy the required certificates. 

Question-109: What all are the recommendation by the Cloudera for the databases?

Answer:  Cloudera manager uses various data bases and data store to store information about Cloudera manager configuration, as well as information such as the health of the system or task progress.  Following are the recommendation from Cloudera

• Choose one of the supported database providers for all of the Cloudera databases.
• It is recommended that install the database is on different hosts then the services, which helps in isolating the potential impacts from failure or resource contention in one or the other. 

Question-110: What information is stored by Cloudera manager server database, and how it helps during restart?

Answer:   Cloudera manager server database contains all the information about services you have configured and their role assignments, all configuration history, commands, users, and running processes.  This is relatively small database < 100MB, but it is important to create backup. Because when you restart processes, the configuration for each of the service is re-deployed using information saved in the Cloudera manager database.  If this information is not available, your cluster cannot start functioning correctly. Has it is highly recommended that you schedule and maintain regular backups of the Cloudera manager database to recover the cluster in the event of the loss of this databases.