|By Srinivasan Sundara Rajan||
|May 15, 2016 03:30 AM EDT||
Cloud Native Applications
As the cloud becomes more of the norm as part of enterprise computing, enterprises now have to deal with the issue of how to ensure that applications effectively use the attributes of cloud. There are monolithic applications from the previous era that are continuing to be migrated to the cloud using a lift and shift approach. With minimal changes, they do benefit from certain attributes of cloud like availability and management, but there is also a new set of application architecture emerging, namely the ‘cloud native applications.'
A cloud native application is designed to take the best advantage of its deployment to a cloud platform. By adopting cloud native applications they improve their agility in the way they build, deploy and manage their applications.
The design patterns of the existing application architecture don't fully support the concept of cloud native applications and require new set of design patterns. There are two important such design patterns that are emerging in the cloud era.
- Data Lake
While both represent two different layers of architecture, there are some ways both are interrelated as explained below.
A lot of information is available about microservices. Over the years multi-tiered and tightly coupled applications have grown and are typically called as "monolithic" applications. However, these "monolithic" applications are not able to take full advantage of cloud platforms and it's difficult to scale them at a component level. Another issue is that these applications tend to have single point of failures due to tight coupling in nature. Microservices address this issue with the design of self-contained services that can be deployed, versioned and scaled independently with the rest of the components of the application. A cloud native approach to application design warrant a microservices design approach where by decoupled federation of services makes up an application. Microservices also provide well-defined interfaces based on industry standard protocols so that they can interface with each other.
In the era of Big Data, a data lake is an enterprise-wide repository of data of any size, type and format. Typically data lake repositories are built on a Hadoop HDFS based file system.
- Unlike a typical data warehouse or an operational data store, data store does not require data validation and cleansing before it gets loaded.
- A data lake does not require Schema On Write, but rather go by Schema On Read. However the data producing applications can use their own schema while pushing data to the data lake.
- A data lakecannot use the traditional JOINS to relate data, but depend on complex processing capability to provide results at low latency.
Issues with Microservices and Usage of DataLake
While there are advantages in using a microservices pattern from a cloud native applications, it does have its limitations. As per the definition of microservices, they have to be self-contained, which means that each microservice has to have a separate data store of its own and can't use a shared database as is typically done in monolithic applications.
Having separate databases per microservice typically makes the data integration and application-wide data access across microservices very difficult. Here is where the concept of data lake comes in handy.
By ensuring that the microservices are managing their own data, if they also write their persistent data to an enterprise-wide data lake, then the data lake can be utilized for centralized data access and integration needs.
The following diagram shows how microservices and data lake are related from enterprise application design perspective.
Microservices and DataLake Support in Major Cloud Platforms
All major cloud platforms started supporting both microservices and data lake as part of their offerings so that the enterprises can build cloud native applications. For example in Microsoft Azure the following support is supported.
- Service Fabric enables you to build and manage scalable and reliable applications composed of microservices running at a Service Fabric cluster.
- Azure Service Fabric offers two high-level frameworks for building services: the Reliable Services API and the Reliable Actors API.
From a data lake perspective, Microsoft Azure provides the following services.
- Azure Data Lake includes all the capabilities required to make it easy for developers, data scientists and analysts to store data of any size, shape and speed and do all types of processing and analytics across platforms
- With Data Lake Analytics, use U-SQL, a query language which blends the declarative nature of SQL with the expressive power of C#.
It may require separate write-ups to go into detail about these two services. But the point is that cloud providers support the design patterns for cloud native applications and organizations can appropriately use them.
- Java EE 7 and Cloud Computing
- Cloud Computing Reference Architecture – Review of the Big Three
- Windows Azure vs VMware vFabric
- PaaS: .NET vs Java EE
- Using Amazon Elastic MapReduce in the Automotive Industry
- Five Factors to Influence Cloud Adoption – The Pros and Cons
- Dynamic Scaling and Elasticity - Windows Azure vs Amazon EC2
- Cloud Analytics - The Big Four Offerings
- Enterprise Java EE PaaS - OpenShift vs Google App Engine for Java
- Challenges and Solutions for the Health Care Industry in Cloud Computing