H13-711_V3.5 Exam Questions

When using the configuration planning tool to plan a Fusioninsight HD V100R002C50 cluster, you can use the configuration planning tool of the V100R2002C60 version for planning.

Flink supports Local mode and Cluster mode deployment, and other mode deployments are not currently supported.

In the Hadoop system, if the backup factor of the HDFS file system is 3, then each time MapReduce runs the Task, it must transfer the file segments to be processed from 3 machines...

In the Spark SQL table, there are often many small files (the size is much smaller than the HDFS block size). In this case, Spark will start more Tasks to process these small files...

Hive in Fusioninsight includes two roles, HiveServer and MetaStore.

As an authentication server center, Kerberos can provide unified authentication services to all services in the cluster and secondary development applications of customers.

server.channels.ch1.transactioncapacity indicates the transaction size, that is, the number of events that the current channel supports transaction processing, which can be set to...

In the Fusioninsight HD system, if dirty data is generated during the running of the Loader job, the status of the Loader job execution result must be failure.

The processing logic of topology is in the bolt.

ElasticSearch is strongly dependent on Zookeeper.

Flink can not only provide real-time computing that supports high throughput and exactly-once semantics, but also batch data processing.

Batch processing of high-value and highly aggregated information and knowledge is the main business appeal of the big data industry

In order to consider performance optimization, it is recommended to deploy LdapServer and KrbServer on the same node in all clusters.

KrbServer can provide kerberos functions for other components, which are used to prevent eavesdropping, prevent replay attacks, and protect data integrity.

Kafka is a distributed message publishing and subscription system. It only forwards messages and does not save messages.

Flink is a computing framework combining batch processing and stream processing. Its core is a stream data processing engine for data classification and parallel computing.

Flink has no restart strategy.

The Streaming of Fusioninsight HD is developed based on the open source Apache Storm, which is a distributed offline computing framework.

The most basic unit of distributed storage in HBase is Region.

The flume architecture shown in the figure below is mainly used to import data from outside the cluster into the cluster.

The main difference between YARN-client and YARN-cluster is the difference between the Application Master process.

Spark divides the stages according to the dependencies of RDDs. The scheduler starts from the end of the DAG graph and traverses the entire dependency chain in reverse. When it enc...

The checkpoint mechanism in Flink continuously draws snapshots of streaming applications, and the status snapshots of streaming applications can only be saved in the HDFS file syst...

In Fusioninsight HD products, a typical Kafka cluster includes several producers, several consumers and a zookeeper cluster?

When using HBase for data reading service in Fusioninsight HD, it is necessary to connect to HMaster.

The process of OpenScanner in HBase will create two different Scanners to read the data of HFile and MemStore. The Scanner corresponding to HFile is StoreFileScanner, and the Scann...

Multiple channels can be configured in the properties.protertises configuration file of Flume to transmit data.

The core concept of MapReduce is to decompose a large computing task to each node of the cluster, make full use of cluster resources, and shorten the running time.

As a distributed message system, Kafka supports online and offline message processing, and provides a java API for other components to connect and use. In the Fusioninsight solutio...

In MapReduce programming, the code should not only describe what to do, but also how to do it.

One of the goals of the data center technology architecture is to be able to solve the disadvantages that traditional data warehouses tend to become data islands and big data platf...

When the size of a Region gradually increases in HBase, it may be pruned.

A KeyValue format in the data file HFile of HBase contains Key, Value, TimeStamp, KeyType, etc.

Flink is a unified computing framework combining batch processing and stream processing. Its core is a stream data processing engine for data distribution and parallel computing.

When Solr creates a Collection, it is recommended to use compositld Router as the routing algorithm, so that the Collection can expand shards.

FLINK uses the checkpoint mechanism to ensure the fault tolerance of the application program.

When using Loader for data import and export, data processing must go through the Reduce phase

In Spark on YARN mode, the node without NodeManager cannot start the executor to execute the Task

In the HDFS mechanism, the NameNode is responsible for managing metadata. Each read request on the client side needs to read metadata information from the metadata disk of the Name...

The Container of the Spark task can run multiple tasks.

The shards index fragmentation of ElasticSearch can break up and distribute index data to different nodes.

Redis judges whether a key exists using the exist command.

The memory size allocated by YARN to Container in the Hadoop system can be set by the parameter yarn.app.mapreduce am.resource.mb

The index data of ElasticSearch can only be saved in the HDFS system.

Commands in Redis are case sensitive.

The Kunpeng computing industry is a full-stack IT infrastructure, industry applications and services built on the basis of Kunpeng processors, including PCs, servers, storage, oper...

The Fusionlnsight HD management node provides a dual-node HA mechanism. After the primary node fails, the primary and backup nodes need to manually implement Failover.

After the Fusionlnisght HD cluster is successfully installed, it is not allowed to modify the configuration of services, roles, and instances.

In the task scheduling of YARN. Once the Application Master applies for resources, it communicates with the corresponding Resource Manager and asks it to start the task

The enhanced features of Zookeeper include adding the audit log of ephemeral node deletion in the audit log.