Overview of Cloud Computing Security Challenges
It all started in 2008 when Google published a paper on map-reduce and then open source started building Hadoop for cluster computing to do more parallel tasks. In doing so one would require large computing resources, with that comes high infrastructure cost. By then Amazon Web Services(AWS) was supporting web hosting and other web-related technologies on their servers from 2006. Amazon grabbed this opportunity to support cloud-based computing resources for industrial and academic purposes. As the demand grew every subsequent year, more players came into the picture. As of now 60% of the cloud computing market is dominated by AWS and the rest by Google’s Google Cloud Platform and Microsoft’s Azure.
Challenges in Cloud Computing
Major challenges in Cloud Computing are as follows:
1. Security and Privacy
When we say security and privacy we are talking about the user data that is stored on cloud service providers(CSP) data centers. A CSP should abide by the rules of not sharing confidential data or any data that matters to the users. The data centers must be secure and privacy of the data should be maintained by a CSP.
2. Availability and Reliability
The data and service from the CSP should be available at all times irrespective of the external condition is the ideal condition. The computing resource should be available for the users and their operability should be reliable. Cloud Computing challenges are basically on the CSP side rather than the user.
This means that if the users want to migrate from one CSP to others, the vendor should not lock-in customer data or services and the migration should be ease. There are different laws over data in different countries.
4. Computing Performance
Cloud Computing is on-demand compute service and supports multitenancy, thus performance should not suffer over the acquisition of new users. The CSP should maintain enough resources to serve all the users and any ad-hoc requests.
5. Service Quality
The service quality should be good and is a major concern of the end-user. The whole ecosystem of Cloud Computing is presented in virtual environments and thus the CSP should give what is promised in terms of service, be it compute resources or customer satisfaction.
CSP’s services should be flexible enough to integrate itself into other platforms and services provided by other CSPs. The data pipeline should be easy to integrate and should drive improved performance.
There are a lot of challenges in Cloud Computing like Big data, long hall transfer, transferring data problems but still, it is the best computing resource available to date.
Types of Cloud Computing delivery model
Now that we know what is cloud computing lets us see what are different services cloud offers. There are three delivery models cloud computing follows:
1. SaaS: Software as a Service
Offers on-demand pay per use of software application to users, unlike licensed software which requires to be bought by the customers. SaaS is a platform-independent service as the end-user is not needed to install the software on the system but can use it from the internet. It is managed fully by the vendor who is supporting the services as only one instance of the software need to be available. Many concurrent users can access the software service on demand and pay as they use it. In this way, computing becomes very cheap and software can be accessed via a browser or lightweight client applications.
Thus, SaaS can be used by end-users. The SaaS products are Google ecosystem of office software, same with Microsoft’s Office 365, and salesforce.
Pros: Universally Accessible from any platform with the internet. No need for computing at the user system, you can work from anywhere. All the computations happen in the cloud. Excellent tool for collaborative working. Multiple users can use the software simultaneously and every user experiences it the same.
Cons: Browser issues may end up bad user experience. Internet performance may dictate overall performance.
2. PaaS: Platform as a Service
This service is made up of a programming language execution environment, an operating system, a web server & a database. Encapsulate the environment where users can build, compile & run their programs without worrying about the underlying infrastructure. In this model, you manage data & the application resources, all other resources are managed by the vendor.
Thus, PaaS is used by developers for deploying their applications, for example, AWS elastic Beanstalk, Heroku and force.com.
Pros: This is a cost-effective rapid method of application development. By this service, developers can easily deploy the application on the web. With this both private and public deployment is possible.
Cons: Developers are sometimes limited to cloud providers languages and tools. Migration issues such as vendor lock-in persist.
3. IaaS: Infrastructure as a Service
It offers the computing architecture & infrastructure, all computing resources but in a virtual environment so that multiple users can access them. Resources include data storage, virtualization, servers & networking. Most vendors are responsible for managing the above four resources. Users will be responsible for handling other resources such as applications, data, runtime & middleware.
Thus, IaaS is basically used by system admins or IT infrastructure team. Examples of IaaS providers are EC2, GoGrid, Rackspace.
Pros: The vendor provides the infrastructure hence enhanced scalability, dynamic workload handling. IaaS is very flexible and works on the same pay per use revenue model.
Cons: There are security issues sometimes. IaaS may also suffer the network and service delays.
Conclusion – Cloud Computing Security Challenges
Cloud Computing with lots of ups and downs is the best engineering service of our generation. As more people come on board for using cloud services, CSPs are providing enhanced services and adoption of CC is increasing.
This is a guide to Cloud Computing Security Challenges. Here we discuss the Introduction, Challenges in Cloud Computing and Types of Cloud Computing delivery model. You can also go through our other suggested articles to learn more–