MongoDB Interview Questions and Answers- Part 2

Answer:

MongoDB is a NoSQL database that has a flexible schema. Transactions are not operations that need to be executed for each writes operation as they incur a hefty performance cost over the single document writes. With a denormalized data model and document-based structure in MongoDB, there is a minimal need for transactions. Since MongoDB enables document embedding, one doesn’t necessarily need to use the transaction to meet a write operation.

Answer:

When running a 32-bit build of MongoDB, the total storage size for a server, including indexes and data, is 2GB. So, you should not deploy MongoDB to production on 32-bit systems. Instead, run a 64-bit build of MongoDB, as virtually it provides no limit to storage size.

Answer:

In accordance with the official MongoDB documentation, the covered query has the following characteristics:

• All fields in the covered query are a part of the index.
• All fields returned in the covered query are in the same index.

As all the fields in the query are part of the index, MongoDB matches a query’s conditions and returns the result by using the same index without looking inside the documents. Since indexes are present in the RAM, fetching data from the indexes is faster than fetching data through scanning documents.

Answer:

A covered query makes the query execution faster as indexes are stored in the RAM or sequentially located on the disk. The covered query has all the fields in the index, so MongoDB can match a query condition and return the result fields using the same index without looking inside the documents.

Answer:

Replica set has many data-bearing nodes and optionally an arbiter node. Among the data-bearing nodes, only one member is deemed primary node, while the other nodes are deemed secondary nodes. The members of the replica set are primary and secondary. You can maintain an arbiter as part of the replica set. Arbiters don’t keep any copy of data, but arbiters play a vital role in the elections that elect a primary if the current primary is not available.

• Primary- It is the only member in a replica set that receives the write operations. MongoDB applies the write operations on a primary, then records operations on the primary’s oplog. Thereby, secondary members replicate that log and apply the operations to the datasets.
• Secondary- A secondary maintains a clone or copy of the primary’s datasets. To replicate data, the secondary applies operations from a primary’s oplog to its dataset in an asynchronous manner. A replica set can have multiple secondaries.

Answer:

All elements of CRUD operations are explained below:

• Create operation: Create or insert operation is used to add new documents to a collection. In case the collection does not exist, it will create one.
  The command to insert or create a document on a collection –db.collection.insert()
• Read operation– As its name signifies, it reads the documents from a collection. The process of reading documents takes place by executing a query. The command to read a document is – db.collection.find()
• Update operation– It is used to modify the existing document. The command to updates a document is – db.collection.update()
• Delete operation– It erases or removes the document from a collection. The command to perform the delete operation is – db.collection.remove()

Answer:

MongoDB doesn’t follow a file system fragmentation. It pre-allocates the data files to reserve space when setting up the server. Therefore, the MongoDB data files are large in size.

Answer:

MongoDB lets users mix and match different storage engines within one MongoDB cluster. It provides a simple, flexible, and reliable approach to support diverse workloads. With MongoDB’s flexible storage architecture, a database automatically manages the data movement between the storage engine technologies with the native replication. It reduces operational complexity as compared to running distinctive database technologies. MongoDB uses two main storage engines, i.e., WiredTiger and MMAPv1.

Answer:

MongoDB helps multiple clients to read and write the same data at the same time. To ensure consistency, it uses locking with other concurrency control measures to prevent different clients from modifying/amending the same data simultaneously.

Answer:

The following are some differences between Redis and MongoDB:

Answer:

1. • MongoDB is much faster compared to CouchDB, but CouchDB is safer than MongoDB.
   • Triggers aren’t available in MongoDB, whereas triggers are available in CouchDB.
   • MongoDB serializes JSON data to BSON; however, CouchDB does not store data in the JSON format.

2. S.NO.	MongoDB	CouchDB
   1	It was developed by MongoDB, Inc in the year 2009.	It was developed and released by Apache Software Foundation in 2005.
   2	MongoDB is based on the C++ programming language.	CouchDB is based on the Erlang programming language.
   3	It is the most popular document store available for deployment on self-managed infrastructure and fully-managed cloud services.	CouchDB is a native JSON document store inspired by Lotus Notes. It is scalable from the globally distributed server clusters to the mobile phones.
   4	MongoDB doesn’t have any secondary database models.	CouchDB has a document store as the secondary database model.
   5	Server OS for MongoDB is Windows, Linux, OS X, and Solaris.	Server OS for CouchDB is BSD, Android, Linux, Solaris, OS X, and Windows.
   6	MongoDB supports predefined data types like integer, string, decimal, double, Boolean, or date.	CouchDB does not support predefined datatypes.
   7	It supports read-only SQL queries through the MongoDB Connector for BI.	It does not support SQL query language.
   8	It supports a single replication method of Master-master replication.	It supports two replication methods- Master-slave replication and Master-master replication.
   9	MongoDB supports in-memory capabilities.	CouchDB does not have in-memory capabilities.
   10	It provides support for multi-document ACID transactions with snapshot isolation.	It doesn’t support ensuring data integrity after the non-atomic data manipulations.

Answer:

• Cassandra uses tables and columns to store data, while MongoDB stores data in JSON-like documents
• Cassandra does not entirely support secondary indexes, but MongoDB mainly relies on indexes to fetch data.
• Cassandra has its (CQL) query language, whereas MongoDB supports popular third-party programming languages, like Java and Python.
• Cassandra depends on third-party tools for aggregation, but MongoDB has a built-in aggregation framework.
• Cassandra utilities distributed architecture, that makes it highly available, while MongoDB depends on a master-slave architecture that gives it a lower fault tolerance.
• Cassandra is best suited for MySQL-style databases with higher scalability, while MongoDB is best suited for storing unstructured dat.

Answer:

Yes, an object will be deleted from the database if you remove the object attribute so, it is essential to eliminate the attribute and then save the object again.

Answer:

Below are the main features of MongoDB:

• Queries: It supports document-based and ad-hoc queries.
• Index Support: Any field in a document can be indexed.
• Replication: MongoDB supports Master-Slave replication. It uses native applications to maintain multiple data copies. Preventing database downtime is a replica set’s features as it has the self-healing shard.
• Multiple Servers: MongoDB database can run over various servers. Data is replicated to foolproof the system during hardware failure.
• Auto-sharding: Distributing data across multiple physical partitions is called shards. Due to auto-sharding, MongoDB has an automatic load balancing feature.
• MapReduce: MongoDB supports flexible aggregation tools and MapReduce.
• Failure Handling: In MongoDB, it is easy to deal with failures. Since there are large numbers of replicas in MongoDB, it gives out high data availability and increases protection against database downtimes such as data center failures, rack failures, network partitions, or multiple machine failures.
• GridFS: MongoDB can store files of any size without complicating the stack. GridFS divides files into small parts and stores them as separate documents.
• Schema-less Database: MongoDB is a schema-less database written in the C++ programming language.
• Document-oriented Storage: It uses a BSON format instead of a JSON-like format.
• Procedures: MongoDB JavaScript works well as a database that uses the language rather than procedures.

Answer:

Aggregation process data records and return computed results. The aggregation operations group values from various documents, and perform different operations on the grouped data to produce a single output. MongoDB offers three ways to perform aggregation:

• Aggregation pipeline,
• Map-reduce function,
• Single-purpose aggregation commands and methods

Answer:

Replication refers to a process of data synchronization across multiple servers. It increases data availability and redundancy. With multiple data copies on different database servers, replication protects the database from the loss of a single server. Replication also enables you to recover from service interruptions and hardware failures.

Answer:

Data in MongoDB is stored in BSON documents-JSON-style data structures. Documents consist of one or more fields, and each field contains a value of a specific data type, including binary data, arrays, and sub-documents. Documents that share a similar structure are organized as collections. Think of documents as analogous to rows in the relational database, fields as columns, and collections similar to tables.

Following are the advantages of using documents:

• Documents correspond to the native data types in different programming languages.
• Embedded arrays and documents reduce the need for expensive joins.
• The dynamic schema supports fluent polymorphism.

Answer:

Embedded documents represent the relationship between data written inside a body. The documents are received, while the related data body is small.

Answer:

The application layer encryption refers to a data-security solution that encrypts almost any type of data passing through the application. When encryption happens at this level, data is encrypted across different layers such as files, disk, and databases. The application layer encryption improves security by decreasing potential attack vectors. Another benefit of application encryption is that it enables organizations to secure sensitive data before storing it into the database or cloud environments since it encrypts specific fields or documents at the application layer.

Answer:

The use of encryption for data in transit and storage media is called storage encryption. Data is encrypted when it passes to the storage devices like tape drives, individual hard disks, arrays, or libraries that contain data.  Using storage encryption with file encryption and a database can offset the risk of losing your data. Similar to network encryption, storage encryption is a blunt instrument that protects all the data on every disk or tape irrespective of the data type or sensitivity of the data. It is an effective way to ensure data is protected and safe. However, it is even more secure to encrypt data in databases at every individual file or column.