Wednesday, 17 November 2021

Top 20 Apache NiFi Interview Questions & Answers

  

Ques: 1). Is there a functional overlap between NiFi and Kafka?

Answer: 

This is a pretty typical question, and the situation is actually extremely complementary. When you have a large number of customers drawing from the same topic, a Kafka broker gives very low latency. However, Kafka isn't built to tackle dataflow problems - imagine data prioritisation and enrichment — Kafka isn't built for that. Furthermore, unlike NIFI, which can handle messages of any size, Kafka prefers smaller messages in the KB to MB range, whereas NiFi can accept files up to GB per file or more. NiFi is an add-on to Kafka that solves all of Kafka's dataflow issues.

 

Ques: 2). What is Apache NiFi, and how does it work?

Answer: 

Apache NiFi is a dataflow automation and enterprise integration solution that allows you to send, receive, route, alter, and modify data as needed, all while being automated and configurable. NiFi can connect many information systems as well as several types of sources and destinations such as HTTP, FTP, HDFS, File System, and various databases.

 

Ques: 3). Is NiFi a viable alternative to ETL and batch processing?

Answer: 

For certain use situations, NiFi can likely replace ETL, and it can also be utilised for batch processing. However, the type of processing/transformation required by the use case should be considered. Flow Files are used in NiFi to define the events, objects, and data that pass through the flow. While NiFi allows you to perform any transformation per Flow File, you shouldn't use it to combine Flow Files together based on a common column or perform certain sorts of windowing aggregations. Cloudera advises utilising extra solutions in this situation.

The ideal choice in a streaming use scenario is to have the records transmitted to one or more Kafka topics utilising NiFi's record processors. Based on our acquisition of Eventador, you can then have Flink execute any of the processing you want on this data (joining streams or doing windowing operations) using Continuous SQL.

NiFi would be treated as an ELT rather than an ETL in a batch use scenario (E = extract, T = transform, L = load). NiFi would collect the various datasets, do the necessary transformations (schema validation, format transformation, data cleansing, and so on) on each dataset, and then transmit the datasets to a Hive-powered data warehouse. Once the data is sent there, NiFi could trigger a Hive query to perform the joint operation.

 

Ques: 4). Is Nifi a Master-Server Architecture?

Answer: 

No, the 0-master philosophy has been considered since NiFi 1.0. In addition, each node in the NiFi cluster is identical. The Zookeeper is in charge of the NiFi cluster. ZooKeeper chooses a single node to serve as the Cluster Coordinator, and ZooKeeper handles failover for you. The Cluster Coordinator receives heartbeat and status information from all cluster nodes. The Cluster Coordinator is in charge of detaching and reconnecting nodes in the cluster. Every cluster also has one Primary Node, which is chosen by ZooKeeper.

 

Ques: 5). What is the role of Apache NiFi in Big Data Ecosystem?

Answer: 

The main roles Apache NiFi is suitable for in BigData Ecosystem are:

Data acquisition and delivery.

Transformations of data.

Routing data from different source to destination.

Event processing.

End to end provenance.

Edge intelligence and bi-directional communication.

 

Ques: 6). What are the component of flowfile?

Answer: 

There are two sections to a FlowFile:

Content: The content is a stream of bytes that transports from source to destination and contains a pointer to the actual data being processed in the dataflow. Keep in mind that the flowfile is merely a link to the content data, not the data itself. The actual content will be stored in NiFi's Content Repository.

Attributes: The attributes are key-value pairs that are associated with the data and serve as the flowfile's metadata. These characteristics are typically used to store values that give meaning to the data. Filename, UUID, and other properties are examples. MIME Type, Flowfile creating time etc.

 

Ques: 7). What exactly is the distinction between MiNiFi and NiFi?

Answer: 

Agents called MiNiFi are used to collect data from sensors and devices in remote areas. The purpose is to assist with data collection's "initial mile" and to obtain data as close to its source as possible.

These devices can include servers, workstations, and laptops, as well as sensors, self-driving cars, factory machinery, and other devices where you want to collect specialised data using MiNiFi's NiFi features. Before transferring data to a destination, the ability to filter, select, and triage it.

The objective of MiNiFi is to manage this entire process at scale with Edge Flow Manager so the Operations or IT teams can deploy different flow definitions and collect any data as the business requires. Here are some details to consider:

To move data around or collect data from well-known external systems like databases, object stores, and so on, NiFi is designed to be centrally situated, usually in a data centre or in the cloud. In a hybrid cloud architecture, NiFi should be viewed as a gateway for moving data back and forth between diverse environments.

MiNiFi connects to a host, does some processing and logic, and only distributes the data you care about to external data distribution platforms. Of course, such systems can be NiFi, but they can also be MQTT brokers, cloud provider services, and so on. MiNiFi also enables use scenarios where network capacity is constrained and data volume transferred over the network must be reduced.

MiNiFi is available in two flavours: C++ and Java. The MiNiFi C++ option has a modest footprint (a few MBs of memory, a small CPU), but a limited number of processors. The MiNiFi Java option is a single-node lightweight version of NiFi that lacks the user interface and clustering capabilities. It does, however, necessitate the presence of Java on the host.

 

Ques: 8). Will we be able to arrange the flow to automotive management after the coordinator is in place?

Answer: 

As Apache NiFi is designed to work on the idea of continuous streaming, the processors are already set for eternity twist by default. Unless we opt to handle a processor without assistance, for example, on an hourly or daily basis today. Apache NiFi, on the other hand, isn't supposed to be a job-oriented matter. When we put a processor in the bureau, it operates all of the time.

 

Ques: 9). What are the main features of NiFi?

Answer: 

The main features of Apache NiFi are.

Highly Configurable: Apache NiFi is highly flexible in configurations and allows us to decide what kind of configuration we want. For example, some of the possibilities are.

Loss tolerant cs Guaranteed delivery

Low latency vs High throughput

Dynamic prioritization

Flow can be modified at runtime

Back pressure

Designed for extension:We can build our own processors and controllers etc.

Secure:

SSL, SSH, HTTPS, encrypted content etc.

Multi-tenant authorization and internal authorization/policy management

 

Ques: 10). Is there a NiFi connector for any RDBMS database?

Answer: 

Yes, different processors included in NiFi can be used to communicate with RDBMS in various ways. For example, "ExecuteSQL" lets you issue a SQL SELECT statement to a configured JDBC connection to retrieve rows from a database; "QueryDatabaseTable" lets you incrementally fetch from a DB table; and "GenerateTableFetch" lets you not only incrementally fetch the records, but also against source table partitions.

 

Ques: 11). What is the best way to expose REST API for real-time data collection at scale?

Answer: 

Our customer utilises NiFi to expose a REST API allowing data to be sent to a destination from external sources. HTTP is the most widely used protocol.

If you want to ingest data, you'll utilise the ListenHTTP processor in NIFi, which you may configure to listen to a certain port for HTTP requests and deliver any data to.

Look at the HandleHTTPRequest and HandleHTTPResponse processors if you wish to implement a web service with NiFi. You will receive an HTTP request from an external client if you use the two processors together. You'll be able to respond to the customer with a customised answer/result based on the data in the request. For example, you can use NiFi to connect to remote systems via HTTP, such as an FTP server. The two processors would be used, and the request would be made over HTTP. When NIFi receives a query, it runs a query on the FTP server to retrieve the file, which is then returned to the client.

NiFi can handle all of these one-of-a-kind needs with ease. In this scenario, NiFi would scale horizontally to meet the needs, and a load balancer would be placed in front of the NiFi instances to distribute the load throughout the cluster's NiFi nodes.

 

Ques: 12). When NiFi pulls data, do the attributes get added to the content (real data)?

Answer: 

You may absolutely add attributes to your FlowFiles at any moment; after all, the purpose of separating metadata from actual data is to allow you to do so. A FlowFile is a representation of an object or a message travelling via NiFi. Each FlowFile has a piece of content, which are the bytes themselves. The properties can then be extracted from the material and stored in memory. You can then use those properties in memory to perform operations without having to touch your content. You can save a lot of IO overhead this way, making the entire flow management procedure much more efficient.

 

Ques: 13). Is it possible for NiFi to link to external sources such as Twitter?

Answer: 

Absolutely. NIFI's architecture is extremely flexible, allowing any developer or user to quickly add a data source connector. We had 170+ processors packaged with the application by default in the previous edition, NIFI 1.0, including the Twitter processor. Every release will very certainly include new processors/extensions in the future.

 

Ques: 14). What's the difference between NiFi and Flume cs Sqoop?

Answer: 

NiFi supports all of Flume's use cases and includes the Flume processor out of the box.

Sqoop's features are also supported by NiFi. GenerateTableFetch, for example, is a processor that performs incremental and concurrent fetches against source table partitions.

At the end of the day, we want to know if we're solving a specific or unique use case. If that's the case, any of the tools will suffice. When we consider several use cases being handled at once, as well as essential flow management features like interactive, real-time command and control with full data provenance, NiFi's benefits will really shine.

 

Ques: 15).What happens to data if NiFi goes down?

Answer: 

As data moves through the system, NiFi stores it in the repository. There are three important repositories:

The flowfile repository.

The content repository.

The provenance reposiroty.

When a processor finishes writing data to a flowfile that is streamed directly to the content repository, it commits the session. This updates the provenance repository to include the events that occurred for that processor, and it also updates the flowfile repository to maintain track of where the file is in the flow. Finally, the flowfile can be moved to the flow's next queue.

NiFi will be able to restart where it left off if it goes down at any point. This, however, overlooks one detail: when we update the repositories, we write the into the repository by default, but the OS frequently caches this. If the OS dies together with NiFi in the event of a failure, the cached data may be lost. If we absolutely want to eliminate caching, we can set the nifi.properties file's repositories to always sync to disc. This, on the other hand, can be a severe impediment to performance. If NiFi goes down, it will have no effect on data because the OS will still be responsible for flushing the cached data to the disc.

 

Ques: 16). What Is The Nifi System's Backpressure?

Answer: 

Occasionally, the producer system outperforms the consumer system. As a result, the messages consumed are slower. As a result, all unprocessed communications (FlowFiles) will be stored in the connection buffer. However, you can set a restriction on the magnitude of the connection backpressure based on the number of FlowFiles or the quantity of the data. If it exceeds a predetermined limit, the link will send back pressure to the producing processor, causing it to stop working. As a result, until the backpressure is removed, no new FlowFiles will be generated.

 

Ques: 17). What Is Bulleting In Nifi And How Does It Help?

Answer: 

If you want to know if a dataflow has any issues. You can look through the logs for anything intriguing, but having notifications appear on the screen is far more convenient. A "Bulletin Indicator" will appear in the top-right-hand corner of the Processor if it logs anything as a WARNING or ERROR.

This sign, which resembles a sticky note, will appear for five minutes after the incident has occurred. By hovering over the bulletin, the user can get information about what happened without having to search through log messages. If in a cluster, the bulletin will also indicate which node in the cluster emitted the bulletin. We can also change the log level at which bulletins will occur in the Settings tab of the Configure dialog for a Processor.

 

Ques: 18). When Nifi pulls data, do the attributes get added to the content (real data)?

Answer: 

You may absolutely add attributes to your FlowFiles at any moment; after all, the purpose of separating metadata from actual data is to allow you to do so. A FlowFile is a representation of an object or a message travelling via NiFi. Each FlowFile has a piece of content, which are the bytes themselves. The properties can then be extracted from the material and stored in memory. You can then use those properties in memory to perform operations without having to touch your content. You can save a lot of IO overhead this way, making the entire flow management procedure much more efficient.

 

Ques: 19). What prioritisation scheme is utilised if no prioritizers are set in a processor?

Answer: 

The default priority strategy is described as "undefined," and it is subject to change. If no prioritizers are specified, the processor will order the data using the Content Claim of the FlowFile. It delivers the most efficient data reading and the highest throughput this way. We've debated changing the default setting to First In First Out, but for now, we're going with what works best.

 

Ques: 20). If no prioritizer square measure set in a very processor, what prioritization plot is used?

Answer: 

The default prioritization theme is claimed to be undefined, and it’s going to regulate from time to era. If no prioritizer square measure set, the processor can kind the info supported the FlowFiles Content Claim. This habit provides the foremost economical reading of the info and therefore the highest output. we’ve got mentioned dynamical the default feels to initial In initial Out, however, straight away it’s primarily based happening for what offers the most effective do its stuff.

This square measure a number of the foremost normally used interview queries vis–vis Apache NiFi. To go surfing a lot of terribly regarding Apache NiFi you’ll be able to check the class Apache NiFi and entertain reach purchase the newssheet for a lot of connected articles.

No comments:

Post a Comment