Bigtable in cloud computing 

Bigtable in cloud computing 

Big table is created to handle large amount of structured data. This data is associated with company’s internet search and Web service operations. This data is column oriented and distributed by Google Inc. Bigtable was designed to cater to the needs of applications that require massive scalability. 

Bigtable in cloud computing easy proprietary data storage system. It is built on file system, chubby lock service and few other technologies. Public version of bigtable was made available to public on May 6, 2015. Bigtable in cloud computing is used by many applications such as google maps, Google Earth, YouTube and gmail. Bigtable assist in scalability and offers better control of performance characteristics. 

Bigtable is an internal Google database system. It’s started an entire industry known as NoSQL. Bigtable in cloud computing has wab indexes behind its search engine. Bigtable is a database that delivers real-time access to petabytes of data. Google uses big table to power its core services.

Bigtable is not a good choice for every application or it is powerful. Google has over billion users.  Thus it builds applications which caters to the needs of this much bigger population and it focuses on organising information. Big table does not only serve the Google applications but it is also meant for different types of services. 

Uses of bigtable in cloud computing 

Bigtable is used for fast access to big data. Bigtable has the characteristics which enable user to decide if it’s a perfect fit for particular application. Big table in cloud computing has high overhead. Does it is suitable for data of size more than 1 TB. Bigtable’s performance gets affected if data elements of more than 10 MB size are stored individually. Cloud storage is a better option for the unstructured objects such as video files which are larger in size. 

Big table is not a relational database. It does not support multirow transactions or SQL. That is my bigtable is answerable for a wide range of applications. Most of all it is and suitable for online transaction processing. Bigtable In cloud computing is designed to store key or value pairs if the user wants to store data with more structure then you need to switch to another database. 

After analysing all the limitations it seems like bigtable isn’t a suitable option for all occasions. But there are certain situations where it proves to be most useful. MapReduce operations is the most common and basic function of big data. Bigtable has a very high throughput. It has a great scalability and that is why it is an excellent storage option. It also offers a great assistance in real time analytics. That means if an application needs to perform analysis while the events are happening then bigtable is the most suitable option. Along with real time analysis ,bigtable works well in case of financial services and internet of things.

Limitations of  bigtable in cloud computing 

Bigtable lacks relational database capabilities. But bigtable has the ability to scale a better than traditional databases. That is why Google came up with a solution to merge these two things together. 

Following are the changes that were made to bigtable in order to support relational database Capabilities.

 – software was added that supported more complex data than simple value pairs. 

 – Rather than putting just one primary index software added secondary indices. 

 – A query language which was like SQL was added to bigtable.

 – It also added properties like atomicity, consistency, isolation and durability.

That is why bigtable gives you best of both worlds. It provides customers with massive scalability and consistency. Additional capabilities are available after paying a certain price. It is Google‘s most expensive database service.

Since its inception bigtable was intended to be used along with petabytes of data. Bigtable uses is simple data model. It is designed to be deployed and clustered systems. It is sparsed, distributed, persistent multi dimensional sorted map.

With the help of row key and indexing of the map, Data is assembled. Data is arranged according to row, Column and time stamps. High capacity is achieved with the help of compression algorithms.

Google maintains its software as a proprietary, in-house technology. But bigtable what is disclosed to the public in a technical paper which was presented at USENIX symposium on operating systems and design implementation in 2006. The thorough description that was provided by Google has allowed other organisations to create bigtable derivatives. 

Let us look at the architecture of bigtable

It consist of three different servers which are master servers, tablet servers and lock servers. System is designed in such a way that clients do not get to communicate with Master. A free tablet server is identified and the excessive load is sent to it. In order to make tablet servers hold on to their locks, master server communicates with tablet server. 

Following are the other examples of bigtable in cloud computing


Java – simple-CLI, hello world, import HBase sequence files, dataproc word count using map or reduce, GAE flexible-hello world and GAE J8 Std-Hello World.

Dataflow- connector-examples, Pardo-hello world and data flow-coinbase

Go – cbt doc, hello world, usercount and search

Python- hello world and hello world happy base

Leave a Reply

Your email address will not be published. Required fields are marked *