Hbase architecture:The Hbasearchitecture consists of four components HMaster HRegionserver HRegions ZookeeperFig 1: Hbase ArchitectureThe functionalitiesof each one is as follows:HMaster is themaster server that acts like a supervisor for all the region servers and thisacts like a interface for all metadata changes, this also runs severalbackground threads.
HMaster runs on NameNode. HMaster has the features like failoverto handle the load over nodes present in the cluster and controlling loadbalancing. It is responsible to assign regions to region servers, it distributesservices to different region servers and provide admin performances it plays crucialrole in terms of performance and maintaining nodes in the cluster. HMastertakes care of any schema and Metadata operations.HRegionserver isresponsible for managing and serving regions and data that is present indistributed cluster.
it server takes care of region servers, whenever RegionServer receives writes and read requests from the client, it assigns request toa specific region. The region servers run on Data Nodes presentin the Hadoop cluster, where HRegion is basically region server implementation.1 HRegionsare made of mainly two components Hfile and Memstore. They are buildingelements of the whole cluster of the distributed tables which consists ofcolumn families.
It contains multiple stores, one for each column family. Zookeeperis a centralized monitoring server which provides distributed synchronization andapplicationsrunning across the cluster that coordination services between nodes. When theclient needs to contact regions it has to go to Zookeeper first. It is used totrack network partitions and server failure. Zookeeper masters the server usabilityto discover available servers in the cluster.2 MongoDB architecture: MongoDB is the database which has beendeveloped to meet the requirements of processing and managing big data.
Hence,in this regard the architecture of the database has been designed and developedaccordingly. MongoDB is a data base of NO SQL, it’s motto is to combine therelational databases and the critical capabilities with the innovation ofproviding a consistent, scalable, robust, schema less and flexible database tomeet the requirement of the modern databases. FIG 2: MONGODBArchitectureMongoDB in the Core works with theExtended version of JSON known as Binary JSON (BSON). As the features if BSONare Lightweight, Efficient, Transverable MongoDB send and receive data throughthis.
Not only that it stores data as BSON object due to its efficient andquick encoding and decoding skills.As a new age database system like Hadoop for the numerous databaserelated tasks in the modern applications, the MongoDB can be implemented. Moreover,it has great features that make it very easy to implement and use which makesit very friendly to be implemented and used with a variety of modern devicesand applications. It can be easily implemented for the Internet of Things (IoT)devices as well as for the applications.MongoDB database can beconfigured to store data from any device and receive it and this data can beeasily queried from various applications in a very efficient manner.3 Key Characteristicsof HBase and MongoDB:To make scalingeasier for MongoDB it is moved away from relational database.
It is a schemewith less document-oriented database. MongoDB is written in C++. The basicconcept is for replacing the concept of “row” with much more flexible modelwith embedded document and arrays, this approach makes it possible forrepresenting complex hierarchical relationship with single record. Moreover, itis a schema-free which means documents key is not fixed or pre-definedMongoDB supports complex and powerful query language and high access tomass data it supports BSON data structures for storing complex data types, storesas well as distributes large binary files like videos and images instead ofstoring procedure developers that can use or store Java functions, scripts andvalues on the server side, it supports an easy-to-use protocol to store largefile and file metadata, and it will give fast serial performance for singleclient and use memory mapped file for fast performance.
4 To increase the efficiency thedatabase MongoDB provides aggregationfeature. To overcome the scaling problem in web applications, MongoDB has comeup with Sharding feature which isone of the main key features of MongoDB. Sharding makes the database possibleto reach horizontal scalability. Furthermore, MongoDB is able to support ad hoc queries by using unique querylanguage and indexing BSON documents.
The other key features of Mongo Db are capped collections, replication and MongoDBManagement Service (MMS) InHBase by adding the coprocessors it reduces the latency involved in process by altogethereliminating it. It can be nicely integrated to Hadoop which makes it moreflexible which is basically high-Performance, Column-Oriented, Distributed,Multi-Dimensional and High-Availability storage system.5The other important keyfeatures of HBase are HBase is linearlyscalable which means it can bescalable depending up on the requirement, it has automatic failure support that means whenever there is a failure of anode it has the feature of automatic recovery, it provides consistent read and writes.Whenever there is a need to write heavy applications HBaseis preferred, as HBase is known for fast access of data it is the most seducingand outlining factor for it to reach out to lot of companies. It is evidentthat because if these features companies like Twitter, Facebook, Yahoo, andAdobe use HBase internally.