Which of the following is the wrong statement about ZooKeeper?
- A . If ZooKeeper breaks during the synchronization message, after the fault is recovered, it can continue to synchronize according to the transmission status before the failure, that is, support the resume of the breakpoint.
- B . ZooKeeper uses a custom atomic message protocol to ensure the consistency of node data throughout the system.
- C . The ZooKeeper cluster elects the Leader role at startup.
- D . After receiving the data change request, the Leader node writes the disk first and then writes the memory.
Which of the following does not belong to the characteristics of the MapReduce component in Hadoop?
- A . Easy to program
- B . Good scalability
- C . Real-time calculation
- D . High fault tolerance
In the HDFS system, the data block is stored in multiple copies. When the data block on a node is lost, a copy of the data needs to be manually copied to the node.
- A . True
- B . False
Which of the following is the wrong description about Flink barrier?
- A . The barrier is periodically inserted into the data stream and flows as part of the data stream.
- B . The barrier is the core of the Flink snapshot.
- C . A barrier separates the data of this cycle snapshot from the data of the next cycle snapshot.
- D . When the barrier is inserted, the data stream is temporarily blocked.
Which of the following application scenarios generate data based on unstructured data? (Multiple Choice)
- A . CRM customer relationship management business
- B . Seismic surveying and mapping analysis business
- C . Taobao online payment service
- D . Financial instrument imaging business
Which of the following statement about the BloomFilter feature of HBase is incorrect?
- A . Used to filter data
- B . Used to optimize the performance of random reads
- C . Will increase the storage consumption
- D . Can accurately determine that a piece of data does not exist
In the Flink technical architecture, () is the computational engine for stream processing and batch processing.
- A . Standalone
- B . Runtime
- C . DataStream
- D . FlinkCore
The checkpoint mechanism in Flink continuously draws a snapshot of the stream application. The state snapshot of the stream application can only be saved in the HDFS file system.
- A . True
- B . False
In the Fusionlnsight HD platform, which components support table and column encryption? (Multiple Choice)
- A . HDFS
- B . HBase
- C . Hive
- D . Flink
All nodes of ZooKeeper can process read requests.
- A . True
- B . False
Which of the following is the incorrect statement about the description of normal and external tables in Hive?
- A . Create a normal table by default
- B . The external table essentially associates the file path on the existing HDFS with the table.
- C . Metadata and data are deleted at the same time when deleting a normal table
- D . When deleting an external table, only the external table data is deleted without deleting the metadata.
Which of the following description of ZooKeeper feature is wrong?
- A . Updates sent by the client are applied in the order in which they were sent.
- B . A message is received by more than half of the servers and it can be successfully written to disk.
- C . Message updates can only succeed or fail with no intermediate status.
- D . The number of ZooKeeper nodes must be an odd number.
Regarding the comparison between Hive and traditional data warehouse, which of the following description is wrong?
- A . Hive is based on HDFS storage. In theory, the amount of storage can be expanded indefinitely, while the traditional data warehouse storage capacity has an upper limit.
- B . Because Hive is based on a big data platform, query efficiency is faster than traditional data warehouses.
- C . Hive metadata storage is independent of data storage, which decouples metadata and data with high flexibility, while traditional data warehouse data has a single application and low flexibility.
- D . Because Hive’s data is stored in HDFS, it can guarantee high data tolerance and high reliability.
SparkStreaming has higher real-time performance than Storm.
- A . True
- B . False
In the Kafka cluster, what is the role of the Kafka server?
- A . Producer
- B . Consumer
- C . ZooKeeper
- D . Broker
Which of the following description about the reliability of the Fusionlnsight HD system is correct? (Multiple Choice)
- A . There are no single points of failure for all components.
- B . All management nodes HA.
- C . Software and hardware health monitoring.
- D . Network plane isolation.
What processes are included in the HBase service of Fusionlnsight HD? (Multiple choice)
- A . HMaster
- B . Slave
- C . HRegionServer
- D . Data Node
The description of HBase Region Split splitting process, which is wrong?
- A . The split process does not really separate the files, just create a reference file.
- B . In order to reduce the size of the data in the Region, Split splits a Region into two Regions.
- C . The table will be suspended duringSplit.
- D . The Region that was split during the split process will suspend the service.
After the Loader job is submitted to YARN in Fusionlnsight HD, the job cannot be stopped manually.
- A . True
- B . False
When using HBase for data read service in Fusionlnsight HD, you need to connect to HMaster.
- A . True
- B . False
Fusionlnsight Manager supports multi-tenant unified management.
- A . True
- B . False
ZooKeeper nodes use ACL to control access policies. ACL can be applied to any znode node in the cluster.
- A . True
- B . False
The () interface in Flink is used for stream data processing, and the () interface is used for batch processing.
- A . Stream API, Batch API
- B . DataStream API, Data Batch API
- C . Data Batch API, DataStream API
- D . Batch API, Stream API
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 Task to process these small files. When there is a Shuffle operation in the SQL logic, will greatly increase the number of hash buckets, which will seriously affect performance.
- A . True
- B . False
The following description of the Flink window, which is wrong? (Multiple choice)
- A . The scrolling windows do not overlap in time.
- B . The scrolling windows overlap in time.
- C . There is an overlap between the time points between the sliding windows.
- D . There is no overlap between the time point between the sliding windows.
The memory size allocated by YARN to the Container in the Hadoop system can be set by the parameter yarn.app.mapreduce.am.resource.mb.
- A . True
- B . False
Which of the following scenarios are not good for Flink components? (Multiple choice)
- A . Batch processing
- B . Iterative calculation
- C . Stream processing
- D . Data storage
Huawei has rich experience in big data solutions in many industries such as finance, public security, railway and government affairs.
- A . True
- B . False
The following statement about ZKFC, which is wrong?
- A . ZKFC (ZKFailoverController) is used as a client of the ZooKeeper cluster to monitor the status information of the NameNode.
- B . The ZKFC process needs to be deployed in the node of the NameNode and the Leader node of the ZooKeeper.
- C . Standby NameNode senses the status of the Active NameNode through ZooKeeper. Once the Active NameNode rock machine,
- D . Standby NameNode will perform the main operation.
- E . HDFS NameNode ZKFC is connected to ZooKeeper, and the host name and other information is saved to ZooKeeper.
Which application scenarios is HBase not suitable for?
- A . Require application scenarios with full ACID characteristics
- B . Massive data application scenarios
- C . High throughput application scenario
- D . Semi-structured data application scenario