Saturday, August 07, 2021

Why Public Cloud?

 


As mentioned in the previously posted article, the public cloud service has become popular among considerably many banks and companies and unavoidably influential enough to worry the Bank of England and the other regulatory bodies of the UK and the EU. Whereas they admit the accessibility and convenience of these public cloud services, these European authorities are worried about sharing their core information of their national economies with these foreign 3rd parties.

Even with their non-disclosure agreement (NDA), the property right of these information data has been no longer exclusive since it is agreed to execute and store them in devices owned by a 3rd party, and these 3rd party service providers are only a handful of the foreign private enterprises nowadays. The European central banks, governments, and other regulatory bodies are thus anxious about the phenomenon that private banks and companies in their nations rely on these foreign public cloud services all the more nowadays.

Well, whenever a brand new technology emerges, these tends to always be a certain degree of an anxiety reaction to it. The public cloud service can be compared to the utility companies supplying electric power from their power plant having emerged since the early 1900s. (Ref. Investopedia, 1st May, 2020) Before these utility companies were established, each individual firm needed to prepare for, maintain, and update their own on-premise power generator while the majority switched to relying the electric power supply from a specialised utility company due to its accessibility and convenience owing to its big scale merit.

※ The following shows the architectural icons and the configuration examples by referring to the imaginary world public cloud architectural design instead of the real world one in order to avoid both the copyright violation and the favouritism to a particular private corporation.

 


As a matter of fact, it is really advantageous for economic agents to increase their share of using the public cloud for their computer based operations and data storage. The public cloud providers bear responsibility for the durability of their infrastructure (servers, storages, and networks) often together with their platform (O/S, middleware, and Runtime) as well as updating their cyber security of them (The platform security depends on the service types, and the users are responsible for the security of applications and data unless using the software as a service (SaaS)). By outsourcing the cyber security update and the maintenance of the infrastructure, possibly with the platform, economic agents can spare more of their time and effort for their daily operations and various projects.


 

In addition to the aforementioned advantage in time saving by outsourcing maintenance and security updates, the scale merit of the public cloud is a remarkable advantage. This scale merit is the most critical key to optimise the net benefit of using the public cloud. Because of using the physical infrastructure, possibly together with the platform, already prepared in the outsourced data centre, the configuration and its redundancy process are agile and flexible.

The most symbolic function of the public cloud is the scalability of instances launched on servers. ]The on-premise server which is statistically set up to serve for a specific purpose e.g. data base server, mail server, web server, etc. Then, the users have to pay the fixed cost for equipping and maintaining these physically present devices. 

 

 On the other hand, the public cloud launches an instance of various types on a server and dynamically increase and decreases the number of servers while also switching the instance types depending on the user's demand at each period. The scalability of the server-instances enables the users to convert their fixed cost of preparing their server usage to the variable cost. In another word, by switching from their on-premise server to this scalable instance (the virtual server in the cloud is often called "instance" in short) they can switch their payment to a pay-per-usage model which avoids paying for unused computer capacities thanks to these scalable server instances in agile and flexible configuration.

In terms of the capacity control, it does not require users to estimate the capacity to use in advance because the public cloud service automatically scales it to match the supplies capacity with the demanded capacity every scheduled period. In terms of the on-premise environment, it is extremely costly and environmentally unfeasible to increase and decrease server machines and replaces their inside instances every short period (hourly and daily). By contrast, the cloud computing offers such an elasticity enabling flexibility to dynamically increase and decrease the number of instances. 
 

On the other hand, the public cloud launches an instance of various types on a server and dynamically increase and decreases the number of servers while also switching the instance types depending on the user's demand at each period. The scalability of the server-instances enables the users to convert their fixed cost of preparing their server usage to the variable cost. In another word, by switching from their on-premise server to this scalable instance (the virtual server in the cloud is often called "instance" in short) they can switch their payment to a pay-per-usage model which avoids paying for unused computer capacities thanks to these scalable server instances in agile and flexible configuration.

In terms of the capacity control, it does not require users to estimate the capacity to use in advance because the public cloud service automatically scales it to match the supplies capacity with the demanded capacity every scheduled period. In terms of the on-premise environment, it is extremely costly and environmentally unfeasible to increase and decrease server machines and replaces their inside instances every short period (hourly and daily). By contrast, the cloud computing offers such an elasticity enabling flexibility to dynamically increase and decrease the number of instances.
 
This is the example of the configuration based on the simple idea. This example tends to be seen in such as a simple relatively small internet shopping site purely operated via the web.  Majority of functions are located in the virtual private cloud (VPC). The VPC is designed for protecting these programmes from the outside access as though forming the private cloud limiting the outside access.  The load balancing function represented by Symmetric Loader distributes the access loads to Scalable Instances (Redundancy is implemented).  
 

Then, the data of transaction, etc. are gathered in Vertically Scalable Database, most notably installing SQL. Then, the gathered data in Vertically Scalable Database are sent to one Scalable Instance specially assigned for an analysis.  
 
Storage Bucket outside the VPC stores the analysis data from this Scalable Instances and the logs of all the programmes' activities. 
 

The access going out of the VPC requires to set up the point where the connection ends there.  The edge of the VPC is equipped with the gateway connecting between the VPC and outside which is represented by Edge Portal here.

This example is more complex.  This model combines the public cloud with the on-premise configuration.  Like a gaming company, there is always a certain amount of a fixed volume of the loads whilst the total volume is very fluctuating over time periods.  
This type of company tends to manage their staff ID access managements and remain their core functions within their on-premise environment. It needs to directly connect the on-premise environment and the public cloud environment to frequently forward and connect their information, data, etc. between them.  Then, there is Straight Access between the on-premise environment and the public cloud environment.
 
There is the domain name service (DNS) is allocated in order to transform to manage the domains and the IP addresses among these access connections.  The DNS also has a function of balancing the loads between the available paths.  
There is a newly introduced database server represented as Horizontally Scalable Database, most notably NoSQL database.  This database is used for more qualitative data which are not relational like what Vertically Scalable Database which handling more quantitative relational data. 
 

Overall, the public cloud services are really passionate to study and image to use in a real business field, and it will certainly keep growing enough to become non-negligible to mention while discussing about economics and other forms of social science. Even in politics, the controversy on this emerging technology will be a hot topic.  This blog project is going to furthermore focus on the present states and the possible and feasible development of the public cloud from now on.