For example, an online transaction and processing system or a scalable database management system can be upgraded to process more transactions by adding new processors, devices, and storage that can be easily and transparently deployed without shutting them down.

A router protocol is considered scalable with respect to the size of the network, if the size of the necessary router table at each node grows as an asymptotic upper bound (log N), where N is the number of nodes in the network.

On the other hand, a software application is scalable if by increasing the processors where it runs, performance grows proportionally. On the contrary, an application is not scalable if its performance does not "scale" or grow with the increase in processors.

Another example is: at CoE = 4.

Capacidad del sistema informático de cambiar su tamaño o configuración para adaptarse a las circunstancias cambiantes.

Vertical scalability

A system scales vertically, or upward, when adding more resources to a particular node of the system improves the system as a whole. For example, adding memory or a faster hard drive to a computer can improve overall system performance. Another very basic example within the digital economy would be to increase the number of CPUs that our website server has.[4]​.

This is the simplest system, because as we have said, it means growing the hardware of one of the nodes, that is, increasing the hardware with a more powerful one, such as a hard drive, memory, processor, etc. but it can also be the complete migration of the hardware for a more powerful one. The effort of this growth is minimal, since it has no repercussions on the software, since it will only support and migrate the systems to the new hardware.

Although we have seen that it is a simple process, the reality is that this type of escalation contains some negative aspects. This is because growth is directly linked to hardware, and this; sooner or later it reaches its limit. There will come a time when we will have the best hard drive, the best processor, the best memory and we will no longer be able to grow any more. We could buy the next model of servers, but the cost of this operation would be enormous and the performance would only improve a little, meaning it would not be worth it. Furthermore, we would still have the same problem next year and therefore it would not be an effective solution. However, at a structural level it does not involve any notable modification, which makes it a good option if the above costs are acceptable.

In any case, this does not mean that the vertical scalability system is bad, since we can combine it with horizontal scaling and thus obtain better results.

Horizontal scalability

A system scales horizontally if adding more nodes to it improves its performance. It is based on maintaining the cost of development and application, adapting at all times to its continuous growth, but it is a more complex system to implement and manage. This type of scalability is based on the modularity of its functionality. For this reason, it is usually made up of a group of equipment that supports the complete functionality. Normally, in horizontal scalability, equipment is added to give more power to the work network.

With an environment of this type, it is logical to think that the processing power is directly proportional to the number of computers on the network. The total processing power is the sum of the physical speed of each computer transferred by the application and data partition spread across the nodes.

If a scalability model based on horizontality is applied, there are no a priori growth limitations. As the main and important defect, this scalability model involves a large modification in the design, which leads to a great deal of design and reimplementation work. If the logic is designed for a single server, the architectural model will likely need to be structured to support this scalability model.

There must be someone in charge of how to create the data partition model on the different computers, this is the developer. Since there are dependencies on access to the application. It is convenient to carry out an analysis of user activity to adjust the operation of the system. With this scalability model, there is a system to which resources can be added almost infinitely and adaptable to the growth of workloads and new users.[5]​.

Scalability counts user growth as a critical factor. It is much easier to design a system with a constant number of users (no matter how high this may be) than to design a system with an increasing and variable number of users. The relative growth of numbers is much more important than absolute numbers.

For example, adding a new computer to a system that balances the load between the old and new computer can improve the performance of the entire system. Another example that we found would be that of an online store. In this case it would be about implementing flexible software that can adjust to the requirements in high peaks of demand and that provides excellent management in terms of performance and optimization.[4]​.

The term scalability is also applied in the business world. The intensive use of information and communication technologies (ICT) by companies generates scalable business models, understanding these as business ideas with high expansion potential, rapid adaptation to international markets and capable of generating greater profits without increasing costs linearly. Even if you're not ready to grow right now, there are things you can do to set yourself up for scalable growth and success: build a solid foundation, focus on scalable solutions, and plan strategically.

An example of companies with these characteristics are startups or emerging companies. These types of companies manage, through the use of technology, to grow quickly and continuously while maintaining flexible structures, adaptable according to market development and with relatively low costs.[7]​.

To begin with, we should understand scalability as a starting point and as an essential path for all entrepreneurs starting with a start-up company. Startups base their business and market their services and/or products through the intensive use of information and communication technologies, while working with a stable business model which allows them to grow quickly and sustainably over time. This strong technological component allows them to scale their business agilely and quickly.

This is why startups have great potential with respect to scalability. This is because startups generally develop innovative business models and operate with lower costs than traditional companies. Furthermore, their main objective is to grow quickly and they operate, as a general rule, with a horizontal organizational chart that allows them to adapt at all times to their continuous growth.

This concept, ICT, refers to the theories, tools and techniques used in the processing and transmission of information: computing, internet and telecommunications.

Technology is the best equipment for any scalable business, since it has established itself as a key tool for the standardization of the company's processes and general strategy.

It is important that we know the types of technology that best adapt to our business model. ICTs are made up of different knowledge, tools, techniques and devices and when embarking on a scalable project we can find in them tailored solutions to get the job done.[4]​.

The application scaling calculation must begin in the development stage, taking into account the requirements. It is important to decide in advance what will be needed to achieve adequate scaling and how it will be done. You should think about factors such as the growth behavior of the application and the limits of infrastructure resources that we can have or assume financially. These decisions will surely condition development and indicate the need for specific architectures (microservices, REST...), so that the software adapts correctly.

Regarding vertical scaling, it is important to know that it is not always possible to increase the current resources of a server. For example, Dedicated Servers have few possibilities for expansion and it is generally necessary to change the entire machine, with the corresponding migration. This, however, does not happen with Cloud Servers, since it is perfectly possible to assign more resources to the server when necessary, saving us the need to change machines.

Cloud Servers also make horizontal scaling much easier, since creating new servers and adding them to the cluster is very quick and easy. The complexity of these configurations does not have to scare us, because thanks to the Arsys Cloudbuilder Next panel it is possible to configure a cluster of servers with load balancing very easily, practically with the click of a button, as you can see in the tutorial When and how to do load balancing on our Cloud Servers.

Generally, vertical scaling is more than adequate for most applications, but when high user growth and therefore a high number of requests are expected in a short space of time, planning a horizontal scaling strategy may be necessary. It should also be considered that sometimes increasing hardware is not viable, due to budget or because the improvement is not relevant enough for an application. In these cases horizontal scaling is much safer, but we can also resort to mixed strategies, such as separating certain services to an additional server, such as the database. In this way, without going into the complexities of configuring a cluster, we can have two machines serving requests, one in charge of the web server part and the other in charge of the database.[8]​.

For example, an online transaction and processing system or a scalable database management system can be upgraded to process more transactions by adding new processors, devices, and storage that can be easily and transparently deployed without shutting them down.

A router protocol is considered scalable with respect to the size of the network, if the size of the necessary router table at each node grows as an asymptotic upper bound (log N), where N is the number of nodes in the network.

On the other hand, a software application is scalable if by increasing the processors where it runs, performance grows proportionally. On the contrary, an application is not scalable if its performance does not "scale" or grow with the increase in processors.

Another example is: at CoE = 4.

Capacidad del sistema informático de cambiar su tamaño o configuración para adaptarse a las circunstancias cambiantes.

Vertical scalability

A system scales vertically, or upward, when adding more resources to a particular node of the system improves the system as a whole. For example, adding memory or a faster hard drive to a computer can improve overall system performance. Another very basic example within the digital economy would be to increase the number of CPUs that our website server has.[4]​.

This is the simplest system, because as we have said, it means growing the hardware of one of the nodes, that is, increasing the hardware with a more powerful one, such as a hard drive, memory, processor, etc. but it can also be the complete migration of the hardware for a more powerful one. The effort of this growth is minimal, since it has no repercussions on the software, since it will only support and migrate the systems to the new hardware.

Although we have seen that it is a simple process, the reality is that this type of escalation contains some negative aspects. This is because growth is directly linked to hardware, and this; sooner or later it reaches its limit. There will come a time when we will have the best hard drive, the best processor, the best memory and we will no longer be able to grow any more. We could buy the next model of servers, but the cost of this operation would be enormous and the performance would only improve a little, meaning it would not be worth it. Furthermore, we would still have the same problem next year and therefore it would not be an effective solution. However, at a structural level it does not involve any notable modification, which makes it a good option if the above costs are acceptable.

In any case, this does not mean that the vertical scalability system is bad, since we can combine it with horizontal scaling and thus obtain better results.

Horizontal scalability

A system scales horizontally if adding more nodes to it improves its performance. It is based on maintaining the cost of development and application, adapting at all times to its continuous growth, but it is a more complex system to implement and manage. This type of scalability is based on the modularity of its functionality. For this reason, it is usually made up of a group of equipment that supports the complete functionality. Normally, in horizontal scalability, equipment is added to give more power to the work network.

With an environment of this type, it is logical to think that the processing power is directly proportional to the number of computers on the network. The total processing power is the sum of the physical speed of each computer transferred by the application and data partition spread across the nodes.

If a scalability model based on horizontality is applied, there are no a priori growth limitations. As the main and important defect, this scalability model involves a large modification in the design, which leads to a great deal of design and reimplementation work. If the logic is designed for a single server, the architectural model will likely need to be structured to support this scalability model.

There must be someone in charge of how to create the data partition model on the different computers, this is the developer. Since there are dependencies on access to the application. It is convenient to carry out an analysis of user activity to adjust the operation of the system. With this scalability model, there is a system to which resources can be added almost infinitely and adaptable to the growth of workloads and new users.[5]​.

Scalability counts user growth as a critical factor. It is much easier to design a system with a constant number of users (no matter how high this may be) than to design a system with an increasing and variable number of users. The relative growth of numbers is much more important than absolute numbers.

For example, adding a new computer to a system that balances the load between the old and new computer can improve the performance of the entire system. Another example that we found would be that of an online store. In this case it would be about implementing flexible software that can adjust to the requirements in high peaks of demand and that provides excellent management in terms of performance and optimization.[4]​.

The term scalability is also applied in the business world. The intensive use of information and communication technologies (ICT) by companies generates scalable business models, understanding these as business ideas with high expansion potential, rapid adaptation to international markets and capable of generating greater profits without increasing costs linearly. Even if you're not ready to grow right now, there are things you can do to set yourself up for scalable growth and success: build a solid foundation, focus on scalable solutions, and plan strategically.

An example of companies with these characteristics are startups or emerging companies. These types of companies manage, through the use of technology, to grow quickly and continuously while maintaining flexible structures, adaptable according to market development and with relatively low costs.[7]​.

To begin with, we should understand scalability as a starting point and as an essential path for all entrepreneurs starting with a start-up company. Startups base their business and market their services and/or products through the intensive use of information and communication technologies, while working with a stable business model which allows them to grow quickly and sustainably over time. This strong technological component allows them to scale their business agilely and quickly.

This is why startups have great potential with respect to scalability. This is because startups generally develop innovative business models and operate with lower costs than traditional companies. Furthermore, their main objective is to grow quickly and they operate, as a general rule, with a horizontal organizational chart that allows them to adapt at all times to their continuous growth.

This concept, ICT, refers to the theories, tools and techniques used in the processing and transmission of information: computing, internet and telecommunications.

Technology is the best equipment for any scalable business, since it has established itself as a key tool for the standardization of the company's processes and general strategy.

It is important that we know the types of technology that best adapt to our business model. ICTs are made up of different knowledge, tools, techniques and devices and when embarking on a scalable project we can find in them tailored solutions to get the job done.[4]​.

The application scaling calculation must begin in the development stage, taking into account the requirements. It is important to decide in advance what will be needed to achieve adequate scaling and how it will be done. You should think about factors such as the growth behavior of the application and the limits of infrastructure resources that we can have or assume financially. These decisions will surely condition development and indicate the need for specific architectures (microservices, REST...), so that the software adapts correctly.

Regarding vertical scaling, it is important to know that it is not always possible to increase the current resources of a server. For example, Dedicated Servers have few possibilities for expansion and it is generally necessary to change the entire machine, with the corresponding migration. This, however, does not happen with Cloud Servers, since it is perfectly possible to assign more resources to the server when necessary, saving us the need to change machines.

Cloud Servers also make horizontal scaling much easier, since creating new servers and adding them to the cluster is very quick and easy. The complexity of these configurations does not have to scare us, because thanks to the Arsys Cloudbuilder Next panel it is possible to configure a cluster of servers with load balancing very easily, practically with the click of a button, as you can see in the tutorial When and how to do load balancing on our Cloud Servers.

Generally, vertical scaling is more than adequate for most applications, but when high user growth and therefore a high number of requests are expected in a short space of time, planning a horizontal scaling strategy may be necessary. It should also be considered that sometimes increasing hardware is not viable, due to budget or because the improvement is not relevant enough for an application. In these cases horizontal scaling is much safer, but we can also resort to mixed strategies, such as separating certain services to an additional server, such as the database. In this way, without going into the complexities of configuring a cluster, we can have two machines serving requests, one in charge of the web server part and the other in charge of the database.[8]​.