Question-1: What is the Cloudera Enterprise?

Answer: Cloudera Enterprise is a combined solution for the Machine Learning, Analytics, Data Engineering etc. Which include the following solutions.

• Data Warehouse
• Data Science
• Data Engineering
• Operational Database
• Run in Cloud, Multi-Cloud or Hybrid Cloud solutions. 

Basically, it’s a combination of following 3 things

-          Open Source CDH (Include Hadoop & its Eco-system)

-          Cloudera Manager (Licensed product from Cloudera )

-          Cloudera Navigator (Licensed Product from Cloudera )

 

Question-2: How does Cloudera Enterprise Differ with the Cloudera Altus?

Answer: Cloudera Altus provide almost the same solution which is provided by Cloudera Enterprise. But in the public cloud like AWS, Azure and Google Cloud. 

 

Question-3: Can you please explain what is the use of Cloudera (SDX) Shared Data Experience?

Answer: Using the Cloudera ’s various solutions like Cloudera Enterprise we can have Data warehouse, data engineering, operational databases workloads altogether on the single platform. In such cases Cloudera Shared Data Experience (SDX) enables these diverse analytic processes to operate against a shared data catalog while having security, governance policies and schema. Even your entire Cloud environment is terminated, it still persists the all the metadata information. 

 

Question-4: Can you please tell me which all components are used as part of Cloudera Data Warehouse solution?

Answer: Currently below 5 major components used to make Cloudera Data Warehouse solution.

• Apache Impala: You can run SQL/BI analytics on the data stored in either of the following

• AWS S3
• Microsoft Azure Data Lake 
• HDFS
• Apache Kudu
• Hive on Spark: This helps in creating faster ETL/ELT solution for BI and Reporting.
• HUE (SQL Development Workbench): At a time 1000’s of SQL developer.
• Workload XM: It is used to analyze the current workload, query analysis and optimization of the cluster resources.
• Cloudera Navigator

 

Question-5:  As part of Cloudera Enterprise Data Science solution, which all are underlined product majorly used or it runs on?

Answer: Currently below 3 major components are used

• Cloudera Data Science Workbench: CDSW provides the on-demand access to Runtime for R, Python, Scala and integration with the Spark framework on CDH. Even for deep learning it supports the GPU accelerated computing and data scientists can use a framework like TensorFlow, MXNet, Keras etc. 
• Apache Spark: Using Spark you can run in-memory processing.
• Cloudera Fast Forward Labs: Using this you can design and execute your enterprise machine learning strategy.
  

Question-6: What is the use of Apache Kudu?

Answer: Kudu is a Hadoop-native storage for fast analytics on fast data. It complements the capabilities of HDFS and HBase. 

Question-7: What is Cloudera CDH?

Answer: It is a distribution from Cloudera for Hadoop and its related projects. CDH is an open source product which include many projects few examples are below.

• Hive
• Impala
• Kudu
• Sentry
• Spark

CDH is considered unified solution for the batch processing, Interactive SQL, interactive search, Machine Learning, statistical computation and role-based access control.

Question-8: Please tell me something about the Apache Hive?

Answer: Hive is a data warehouse solution for reading, writing and managing large datasets in distributed storage like HDFS using Hive Query Language (Almost same as SQL). These queries are converted into a series of jobs which execute on a Hadoop Cluster using either MapReduce or Spark.

 

Question-9: There are many tools available for querying the data, then why to use Hive?

Answer: Hive is a petabyte-scale data warehouse system which is built on the Hadoop platform. And one of the best available choices where you expect high growth of data volume. Hive on either MapReduce or Spark is best suited for batch data preparation or ETL.

Question-10: Can you please give me some use cases where Hive should be used?

Answer: Let’s see few of the below of the use cases

• Suppose you have large ETL Sort and Join jobs to prepare the data for BI users in Impala then schedule such ETL jobs in the Hive. 
• Suppose you have a Job where data transfer or conversion take many hours and possibility of job failure in between then do such activity using Hive, which can help you in recovering and continues where it left.
• Various formats of the data, suppose you are receiving data in various formats then with the Hive SerDe and Variety of UDFs can help in converting data in single format.
  

Question-11: Can Hive metastore used by other Hadoop components?

Answer: Yes, Hive metastore contains the information regarding data stored on HDFS, so that other Hadoop components like Impala can leverage that. Even if you don't have Hive then also this Metastore would be used. 

Question-12: What do you mean by Remote Mode of Metastore?

Answer: Remote mode means metastore should be running in its separate JVM process. And any other process which wanted to get connected with the Metastore for example HiveServer2, HCatalog, Impala etc. should use the Thrift network API.

Question-25: Can you please tell me some benefits of the Apache Kudu?

Answer: Following are the few benefits of the Kudu

• Fast processing of OLAP workloads
• It can be easily integrated with the MapReduce, Spark, Flume & Other Hadoop Components.
• Tight integration with Impala
• Strong but flexible consistency e.g. consistency per request basis.
• Highly performant for running sequential and random workloads simultaneously. 
• Can be managed using Cloudera Manager
• Structured Data Model
• Highly available
  

Question-26: What kind of applications where Kudu best fit?

Answer: There are following things which are difficult to implement on currently available Hadoop Technologies, but Kudu can help

• Reporting application: Where new data must be immediately available for end users.
• Time-series applications: Querying large amount of historic data as well as granular queries on individual entity.
• Predictive Models: Application which uses the predictive models for making real-time decisions, with the periodic refreshes of the predictive models based on historical data.
  

Question-47: Which all are the software distributions are supported by Cloudera Manager?

Answer: Cloudera Manager support two software distribution formats

• Packages: Package is a binary distribution format which contains compiled code, meta-information, dependencies etc. Cloudera Manager uses the native system package manager for each supported OS.
• Parcels: Is a binary distribution format containing the program files, metadata etc.  
  

Parcels are self-contained and installed in a versioned directory, which means that multiple versions of a given parcel can be installed side-by-side. You can then designate one of these installed versions as the active one. With packages, only one package can be installed at a time so there is no distinction between what is installed and what is active. If you want to have Rolling Upgrade enabled then parcels are also required and package does not support the rolling upgrades.