NoSQL Database Comparison – Alibaba Cloud, AWS, Google Cloud, IBM and Microsoft

Data is everywhere around us and we interact with it regularly. Whether you’re checking out the latest model of a smartphone or buying groceries online, you are interacting with data in one way or the other. We have been dealing with data for ages, what has changed now is the scale of data produced and the speed at which it is accessed.

Thanks to digital technologies like cloud, IoT (Internet of things), AI (Artificial Intelligence), machine learning, and more, companies are producing data at an exponential rate. This amount of data collected around the globe is too hefty to process.

According to a report, “People are generating 2.5 quintillion bytes of data each day.”

Here, traditional relational databases like SQL might not offer the required scalability and performance to process large amounts of data. Though relational databases are still relevant, alternative databases like NoSQL come with their own sets of advantages.

This article will help you understand:

.What is NoSQL database?

NoSQL stands for ‘not only SQL’. Big infrastructure providers like Google, Amazon, and Facebook recognized the scalability issue of SQL databases and hence introduced their alternative solutions like Bigtable, DynamoDB, and Cassandra, respectively to meet the changing demands.

NoSQL is an approach to database design. It can accommodate a wide variety of data models. Some of these include key-value, columnar, document, and graph formats. It offers improved scalability, performance, reliability, and consistency as compared to schema-based traditional relational databases.

The NoSQL databases are purpose-built to work with large sets of distributed data. They mostly refer to the databases built in the early 2000s for large-scale clustering of data produced by cloud and web applications.

The need for performance and scalability in cloud-based web applications outweighs the rigid data consistency that traditional relational database management systems provide.

NoSQL database is a form of unstructured storage. They do not have any fixed table structure – one important trait that differentiates them with the common relational databases.

  • NoSQL databases have a flexible schema. There can be different rows having different attributes or structure.
  • They work on a BASE model – Basically Available, Soft state, Eventual consistency.
  • In NoSQL, queries may not always see the latest data. Thus, consistency is guaranteed after some period.

.The advantages of NoSQL databases

  • NoSQL databases have a simpler structure without a schema and are flexible.
  • They are based on key-value points. This means that records are stored and retrieved using a key that is unique to every set of record.
  • NoSQL database can also have column store, key-value graph, document store, object store and other popular data store modes. Thus, they are multi-purpose as well.
  • Open-source NoSQL databases don’t require any expensive licensing fees.
  • They are easily scalable whether you are using an open-source or a proprietary solution. This is because the NoSQL databases can scale horizontally to distribute the load on all nodes. In SQL, this is done by replacing the main host with a higher capacity one, i.e. via vertical scaling.

Now when you know what NoSQL is and what are its advantages, it is time to look at some of the top NoSQL database solutions offered by leading service providers like AWS, Google Cloud, IBM, Alibaba Cloud, and Microsoft. We will also be taking a look at a NoSQL database comparison table for understanding the key differences.

.Comparison between NoSQL database solutions

1. Alibaba Cloud Tablestore

Tablestore is a fully managed NoSQL cloud database service offered by Alibaba Cloud. It can store a large amount of structured and semi-structured data using a variety of data models. Users can use Tablestore database to query and analyse data. Users can also migrate heterogeneous data to this database without any interruptions. With elastic resources and pay-as-you-go billing, it is an efficient and low-cost database management system. It offers high-consistency and service availability with globally spread data centres. Furthermore, distributed architecture and single table auto-scaling makes it highly elastic.

  • It is a fully managed database service. Users can simply focus on business research and development activities instead of worrying about hardware and software presetting, faults, configurations, security, etc.
  • With in-built shards and load balancing, Tablestore can automatically adjust the size of partitions, allowing users to store more data.
  • It creates multiple backups of data and stores them in different server racks.
  • It also offers consistency across three backups. The application can quickly read the data.
Source: Alibaba Cloud
2. Amazon DynamoDB

Amazon DynamoDB is a fast and flexible NoSQL database service that can deliver single-digit millisecond performance at any scale. It is a key-value and document database that is multi-region, fully managed, and durable.

The multi-master database is backed with in-built security, backup, restore, and in-memory caching for internet-scale applications.

  • DynamoDB is built to support the world’s largest-scale applications.
  • Users can build applications with unlimited throughput and storage.
  • The data stored in the database is replicated across multiple AWS regions. This allows local access to data for globally distributed applications.
  • Another great advantage of DynamoDB is that it is serverless. Users have no servers to manage or provision.
  • The database is designed to scale automatically up and down as per the system and capacity requirement.
  • It also supports ACID (Atomicity, Consistency, Isolation, and Durability) transactions – making it an enterprise-ready database.
Reference Architecture of a Weather Application, Source: Amazon Web Services
3. Azure Cosmos DB

Azure CosmosDB is a NoSQL database service by Microsoft that is globally distributed and multi-model. It allows users elastically and independently scales workloads with a click of a button.

Users can also take advantage of fast, single-digit-millisecond data access with the help of APIs like Cassandra, SQL, MongoDB, Gremlin or Tables. It provides comprehensive service level agreements (SLAs) for latency, throughput, availability, and consistency guarantees.

  • With globally spread Azure regions, users can build highly responsive and highly available applications worldwide.
  • It provides 99.999% availability for both write and read actions. It is deeply integrated with Azure Infrastructure and Transparent Multi-Master replication.
  • It offers unprecedented elastic scalability through transparent horizontal partitioning and multi-master replication.
  • Users do not need to deal with index or schema management as the database engine is fully-schema-agnostic.
Source: Microsoft
4. Google Cloud Bigtable

Cloud Bigtable is the fully managed and scalable NoSQL service by Google Cloud. It is best suited for large analytical and operational workloads. It allows users to store terabytes or even petabytes of data. It is ideal for storing large amounts of single-keyed data with very low latency. Cloud Bigtable stores data in scalable tables. These tables are composed of rows and columns. Each of the rows describes a single entity and are indexed by a single row key.

Data stored inside the Cloud Bigtable database is completely secure. The access to the data is controlled by Google Cloud project and the Identity and Access Management (IAM) roles. It also allows users to save a copy of schemas and data as backups.

  • Cloud Bigtable database is designed to scale in direct proportion to the number of machines in a cluster.
  • It can handle upgrades and restart automatically.
  • Users can also increase the size of a Cloud Bigtable cluster for a few hours to manage any large loads.

It is ideal for time-series data, marketing data, financial data, internet of things’ data, and graph data.

IoT use case reference architecture, Source: Google Cloud
5. IBM Cloudant

IBM Cloudant is a fully managed database service designed for hybrid multi-cloud applications. It is built on open-source Apache CouchDB and has a fully compatible API that allows data syncing to any cloud or the edge.

It is a distributed database service that can handle heavy workloads of large, fast-growing web and mobile apps. It is available as an SLA-backed and fully managed IBM Cloud service. Users can also download the service for on-premises installation.

  • It allows users to instantly deploy an instance, create a database, and independently scale.
  • It is ISO 27001, SOC 2 Type 2 compliant and HIPAA ready.
  • With 55+ data centres across the world and globally spread IBM Cloud regions, users can seamlessly distribute data across zones, regions, and cloud providers.
  • The service is compatible with Apache CouchDB, enabling users to access a wide variety of language libraries and rapidly build new applications. Thus, the service boasts of zero vendor lock-in.
AI use case, Source: IBM

Note: The services mentioned above in the NoSQL database comparison have been listed in alphabetical order.

.Tabular Comparison

NoSQL Database Comparison: DynamoDB Vs Bigtable Vs Cloudant Vs Tablestore Vs Azure CosmosDB

Comparison Points DynamoDB Cloud BigTable Cloudant Tablestore Azure CosmosDB
Developed By Amazon Web Services (AWS) Google IBM Alibaba Cloud Microsoft
Primary Database Model
  • Document Store
  • Key-Value store
Wide Column store Document Store Wide Column Store
  • Document store
  • Graph Store
  • Key-Value Store
  • Wide Column Store
Initial Release 2012 2015 2010 2016 2017




Commercial Commercial Commercial Commercial Commercial
Cloud-based Yes Yes Yes Yes Yes
Data Schema Schema-free Schema-free Schema-free Schema-free Schema-free
Server OS Hosted Hosted Hosted Hosted Hosted
Supported Programming Languages*
  • NET
  • Ruby
  • Erlang
  • ColdFusion
  • Groovy
  • Java
  • JavaScript
  • PHP
  • Perl
  • Python
  • C#
  • Go
  • C++
  • Java
  • JavaScript (Node.js)
  • Python
  • C#
  • Java
  • JavaScript
  • Objective-C
  • Ruby
  • PHP
  • Java
  • Python
  • .Net
  • C#
  • Java
  • JavaScript (Node.js)
  • Python
  • MongoDB Client drivers
  • Eventual Consistency
  • Immediate Consistency
  • Immediate Consistency (for single clusters)
  • Eventual Consistency (for two or more replicate clusters)
Eventual Consistency Immediate Consistency
  • Bounded Staleness
  • Consistent Prefix
  • Eventual Consistency
  • Immediate Consistency
  • Session Consistency
Durability Yes Yes Yes Yes Yes
Partitioning Methods Sharding Sharding Sharding Sharding Sharding
Use Cases*
  • Ad Tech
  • Gaming
  • Retail
  • Banking and Finance
  • Media and Entertainment
  • Software and Internet
  • Financial Analysis
  • Internet of Things (IoT)
  • AdTech
  • Web and mobile apps
  • AI solutions
  • IoT apps
  • Social IM
  • Gaming
  • Finance
  • IoT
  • Logistics
  • Mission-critical applications
  • Real-time retail services
  • Real-time IoT device telemetry


Supported Data Types*
  • Scalar: Number, String, Binary, Boolean, and Null
  • Multi-Valued: String set, Number Set, and Binary Set
  • Document: List and Map
Treats all data as raw byte strings for most purposes NA
  • String
  • Integer
  • Double
  • Boolean
  • Binary
Latency Microsecond latency with DynamoDB Accelerator (DAX) Consistent sub-10ms latency NA Low latency Read latency for all consistency levels is guaranteed to be less than 10 milliseconds at the 99th percentile
Replication Automated Global Replication Yes NA NA Transparent multi-master replication
Triggers Yes No Yes No JavaScript
Support for ACID transactions Yes Atomic single-row operations No Atomic single-row operations Yes
Data Encryption Yes Yes (data encrypted at rest) Yes (Data encrypted at rest) NA Yes (Data encrypted at rest)
Backup and Restore On-demand backup and restore Available CouchBackup for snapshot backup and restore Custom backup and restoration Automatic and Online Backups
MapReduce No Yes Yes No Yes (with the help of Hadoop Integration)

Points marked asterisk (*) define an inclusive list in the above NoSQL Database Comparison Table

Suggested Reading: Relational Database Comparison – Alibaba, Amazon, Google, IBM and Microsoft

.Picking the right NoSQL database service – tips

NoSQL database services include a wide and comprehensive set of feature-rich solutions to help you build better applications. However, you should not pick a database just because it offers a lot of features. You need to decide what your application and business needs are. Also, you need to consider factors like vendor-lock in to avoid being stuck with a single service provider.

While all the major NoSQL databases we discussed are popular and enterprise-ready, here are a few things you might want to consider when picking a NoSQL service:

  • Define your database goals: Whether you want to store data as a record; build interactive applications requiring real-time data processing; store data for a backend customer application, etc.
  • Consider the security of data: When you trust a service provider with your data, you must ensure that your data is not compromised and is safe and readily available when required.
  • Consider latency: Latency defines the time taken for a web application to respond to a user’s query. For customer-facing applications, you should consider a database that offers the lowest latency.
  • Consider the hosting choice: You can either go for a self-hosted or a managed database service. Again, it depends on your application requirement. For complex and mission-critical applications, managed services come handy.

We hope our NoSQL Database comparison will help you make the right choice.

Feel free to share your queries and feedback through the comment section below.

Disclaimer: The information contained in this article is for general information purpose only. Price, product and feature information are subject to change. This information has been sourced from the websites and relevant resources available in the public domain of the named vendors on 4th September 2020. DHN makes best endeavors to ensure that the information is accurate and up to date, however, it does not warrant or guarantee that anything written here is 100% accurate, timely, or relevant to the website visitors.


Cloud Cloud News

Veeam expands partnerships with AWS, Azure and IBM Cloud

After landing a $500 million investment from Insight Venture Partners, the leading data management firm Veeam has announced a number of new capabilities for the latest version of Veeam Availability Suite.

Veeam Availability Suite 9.5 Update 4, the latest version, will provide the customers data retention at a lower cost, easier cloud migration and data mobility, and cloud-native backup and protection for AWS. The solution is also going to offer portable cloud-ready licensing, and increased security and data governance. Service providers will now find it easier to deliver Veeam-powered services to customers.

These new capabilities will also be coming to the upcoming Veeam Availability for AWS and Veeam Availability Console v3. The company has expanded its relationships with Amazon Web Services (AWS), Microsoft Azure, IBM Cloud, and over 20,000 service providers.

“Veeam was born and has dominated the backup and data management market in the modern highly virtualized on-premises environments. In the last few years, we have continued our tradition of innovation and evolved to become the leader in Cloud Data Management,” said Ratmir Timashev, co-founder and Executive Vice President of Sales & Marketing at Veeam.

Veeam Availability Platform will now leverage Veeam Availability Suite 9.5 Update 4, Veeam Availability for AWS, Veeam Instance Licensing (VIL), and Veeam Availability Console v3. These major capabilities will provide customers agility, availability, and business acceleration.

“Our latest version of Veeam Availability Platform is one of our most important and anticipated releases to date, providing simple, flexible and reliable solution to help customers migrate to and keep data available in the hybrid cloud regardless of its location,” added Timashev.

The updated Veeam Availability Suite will include cloud tier, cloud mobility, enhanced Veeam DataLabs, intelligent diagnostics, and enhanced Veeam Cloud Connect Replication for Service Providers (VCC-R).

“Today’s announcements reinforce our position as a market leader in Cloud Data Management by expanding our strong relationships with AWS, Microsoft Azure, IBM Cloud, and over 20,000 service providers. We are recognized as one of Forbes 2018 World’s Best 100 Cloud Companies and have been previously recognized twice as Microsoft ISV Partner of the Year – clear proof points of leadership in cloud data management – and this latest launch strengthens our position,” concluded Timashev.

Also read: Veeam makes its hyper-availability solutions available with Lenovo SDI and SAN offerings

The company mentioned that the new and enhanced offerings will further Veeam’s strategy to provide simple, flexible, and reliable solution to customers.

Cloud Cloud News News

IBM brings together Cloud Foundry and Kubernetes with new cloud service

IBM has brought together two of the top open-source technologies of cloud application development – Cloud Foundry and Kubernetes, with its experimental new service called IBM Cloud Foundry Enterprise Environment. 

The Cloud Foundry has been already available on IBM Cloud for a long time now. It powers the applications at companies like Ford Motor. IBM has now included the other leading open-source developer tool, Kubernetes as well, on its cloud.

Kubernetes was open-sourced by Google for automated management of software containers. The containers are emerging as a very important source for cloud applications deployment. The flexibility to move across different types of infrastructure without making any changes in code makes Kubernetes a go-to developer tool to manage cloud application deployments.

Cloud Foundry Enterprise Environment provides a simple web application developer experience, where the developers won’t need to provision any virtual machine, and will have no runtimes or application servers to install.

Developers will have the portability of applications across multiple Cloud Foundry environments, which can help them to roll out code from development environment to production faster than the traditional methods.

Moreover, it will enable automated application lifecycle management, and administrative control over the full environment.

According to IBM, the Cloud Foundry Enterprise Environment will deal with all the Kubernetes stuff that runs underneath and allows the Cloud Foundry developers to innovate on top of that.

“Cloud Foundry leverages the container concept, but provides for a developer experience on top of that. It’s not just screwing around with Kubernetes, but about building applications,” said Don Boulia, general manager for IBM’s Cloud Platform, at Cloud Foundry Summit in Boston.

The new platform has been released as a closed experimental service, and it will come into beta as the company makes improvements to it.

Also read: IBM expands z14 mainframe portfolio to make it a better fit for cloud datacenters

Recently, IBM reported its first -Quarter results and posted a cloud revenue of $17.7 billion over last twelve months, showing a growth of 22% YoY.


Google and IBM to jointly address developer security challenges with Grafeas

Google and IBM recently joined forces to create and open source the Grafeas project, with an aim to provide developers a structured way of auditing and governing the modern software supply chains.

Grafeas provides an open API that collects and aggregates the metadata generated at various stages of software supply chain. The metadata store and enforcement point help in gaining visibility into development environments and in enforcing policies without slowing down the development teams.

IBM has an in-built Vulnerability Advisor into its Container service as a part of DevOps process that scans the container images and detects software package vulnerability and poor software configurations. It further makes a risk assessment for the contained software.

To build a more comprehensive security and governance model, the data can now be combined with other metadata in an open manner using the Grafeas API.

The new project includes the security and governance solutions from Google which will be useful across millions of releases and billions of containers.

The security and governance solutions from Google include- using the immutable infrastructure to establish preventative security postures against persistent advanced threats, building security controls into the software supply chain to protect production deployments, and keeping the system flexible and ensuring interoperability of developer tools around common specifications and open-source software.

Google has also introduced Kristis as an additional component which enables developers creating Kubernetes governance policies on the basis of metadata stored in Grafeas.

Organizations can now store metadata about components from several repositories. Grafeas is hybrid cloud-friendly, and pluggable which helps adding new metadata producers and customers.

Grafeas provides structured metadata schemas, strong access controls, and rich query ability, helping organizations in modern software development environments.

Also read: IBM takes another step towards open source commitment – open sources WebSphere Liberty code

IBM will deliver Grafeas and Kristis as part of IBM Container Service on IBM Cloud. Along with IBM, many well-established organizations including JFrog, Red Hat, Black Duck, Twistlock, Aqua Security, and CoreOS will be contributing to new Grafeas project.

Cloud News

Akamai CDN now available on IBM Cloud

IBM and Akamai Technologies announced the availability of Akamai Content Delivery Network (CDN) on IBM Cloud. The new offering aims to create one of the world’s fastest and most reliable content delivery network.

The new IBM Cloud CDN with Akamai will accelerate and streamline the content delivery including general web delivery, content caching, content purge, ability to view historical metrics for up to 90 days, HTTPS support with wildcard certificates, access control, and more.

“Enterprises are increasingly relying on the cloud to transform and deliver a spectrum of critical business applications to their users,” said Faiyaz Shahpurwala, general manager, IBM Cloud. “By combining the global reach of IBM Cloud with Akamai’s delivery and optimization capabilities, we’re giving businesses the tools they need to innovate in the market and deliver better customer experiences.”

The new service integrates the presence of Akamai in over 1700 networks in 131 countries with global cloud footprint of IBM of 60 cloud data centers across 19 countries. The clients will have the fast access and low latency of requested content for any given region around the world.

IBM and Akamai are long standing partners and by making Akamai technology directly available to IBM cloud customers, the companies will jointly help enterprises in optimizing applications’ performance, speed time to market and improve end user experience.

“The promise of what the cloud can do for business is nearly limitless,” said Rick McConnell, president and general manager, Web Division, Akamai. “At the same time, as an increasing number of business-critical workloads move to the cloud, enterprises seek the assurance of scale, performance and security supporting their applications. This is an exciting time to be part of the cloud ecosystem, and we look forward to further collaboration between Akamai and IBM to help our joint customers achieve their goals.”

Also read: IBM takes another step towards open source commitment – open sources WebSphere Liberty code

The new service is now generally available. The access path for existing users of Verizon Digital Media Edgecast CDN will now be under Network CDN rather than Storage CDN.


IBM increases Datacenters to enable organizations build futuristic applications leveraging IoT & AI

IBM announced the opening of four new datacenters; two in London; one in San Jose, California; and another one in Sydney, Australia.

The announcement came one day after the company’s second quarter 2017 earnings report, which failed to earn good revenues.

With this, IBM now has nearly 60 Datacenters, spread across 19 countries to support the global expansion of its cloud services.

John Considine, General Manager for cloud infrastructure services, IBM said, “IBM operates cloud data centers in nearly every major market around the world, ensuring that our clients can keep their data local for a variety of reasons – including performance, security or regulatory requirements.”

He further added, “We continue to expand our cloud capacity in response to growing demand from clients who require cloud infrastructure and cognitive services to help them compete on a global scale.

Image Credit: IBM

By expanding the cloud footprint, IBM plans to give global organizations access to new facilities that will help them build future oriented IoT (Internet of Things), blockchain and AI (Artificial Intelligence) applications through proper cloud infrastructure facility.

Also Read: AI and cognitive computing – spearheading enterprise digital transformation

With this, clients will also get the flexibility to store and manage data based on their choice.

IBM also says that the new datacenters will help clients meet regulatory compliance standards, as they are one of the first companies to adopt EU’s Data Protection code for CSPs (Cloud Service Providers).

With access to more than 150 APIs, organizations can take help of the cognitive and data computing power technologies which are at the core of these four new datacenters.

A recent report by IDC predicted a $266 Billion worldwide spending on cloud services and infrastructure, which will drive the datacenter industry growth as well.

With this expansion, IBM will be able to strengthen its cloud services by supporting clients’ easy migration to cloud.


AI and cognitive computing – spearheading enterprise digital transformation

Artificial Intelligence, robotics and machine learning concepts are changing the parameters of data extraction and analysis. With the onset of digital transformation, there’s a great need to process huge volumes of data into useful insights.

But with traditional data analytic methods, businesses are not be able to process the entire data, especially that which is in the form of images, videos, and human voice, collectively known as ‘dark data’. To process such kind of data, organizations need cognitive computing.

Originally, the concept of cognitive computing comes from cognitive science which include study of human brain and how it works. Cognitive computing aims to stimulate the human thought processing in a computerized way. It achieves that with the help of various self-learning algorithms, Natural Language Processing(NLP), Machine Learning and Automated Reasoning.
Companies are trying to deliver AI and cognitive based solutions to help organizations deal with various structured and unstructured data sets.

Companies are trying to deliver AI and cognitive based solutions to help organizations deal with various structured and unstructured data sets.

Also Read: Apple adopts the AI move, introduces new machine learning API framework

Companies like Facebook and IBM are rolling out new cognitive and AI solutions for better data analytics. Facebook’s team of researchers from FAIR (Facebook Artificial Intelligence Research), have reportedly introduced new capability to dialog agents – the capability to negotiate. Per them, the dialog agents with differing goals can participate in start-to-finish negotiations with other agents or people to arrive at common decisions or outcomes.

Also Read: New AI server designs to accelerate services of Facebook, Microsoft 

On the other hand, IBM has launched a suite of cognitive computing solutions with an objective to help financial institutions manage their regulatory and fiduciary responsibilities. The suite is powered by Watson which can be deployed on the IBM Cloud. It will help the financial professionals to – understand the regulatory requirement, deliver increased insights into current or potential financial crimes and manage the financial risk with new architectural data approach.

Find more info here.

Organizations need more such cognitive and AI enabled smarter solutions and tools for their digital transformation.

Articles Cloud News News

IBM revenue dip continues in the fifth year

The shares of International Business Machines (IBM) corp. saw a huge decline in revenue for the first time in five quarters. A low demand for company’s traditional IT services has been identified as the reason behind the great fall.

The shares went down by nearly 4.94 percent to $162.00 in the after-hours trading and refused to gain around on Wednesday.

IBM has been shifting its focus towards advanced service portfolio, including cloud-based services, data analytics, security software and artificial intelligence since the demand for its legacy products (hardware and software) came down.

The company was able to grow in the first quarter due to its strategic innovations but failed to remove loopholes from its core operations – like in the technology services and cloud-platforms’ business. IBM missed out on some of the largest deals in the business, and a couple of big clients took their operations in-house per the CFO, Martin Schroeter, IBM. All this resulted in IBM’s revenue coming down and losses accumulating to 2.8 percent in the first quarter to 1.3 percent in the fourth quarter.

Total revenue reportedly fell 2.8% to $18.16 billion, which even missed the analysts’ estimate of $18.39 billion. It also saw a decline of 2% in its infrastructure services.

“I think the frustration lies with the overall miss on revenue,” Bill Kreher – Analyst, Edward Jones, told a news channel. “The Street has given IBM some credit over the last year that the transition is taking shape, so I think that’s where the risk lies execution needs to be strong.”

Revenue of its technology services and cloud services dropped 2.5 % to $8.2 billion. The business accounted for nearly 45 % of the total revenue.The net income of IBM slumped to $1.75 billion from $2.01 billion.