A photo taken in the kitchen ends up \u201cin the cloud\u201d. The phrase suggests a small upward journey, as though the file becomes lighter until it disappears. In reality, it begins a journey through equipment, fibre-optic cables and servers. Somewhere, inside a building, a storage device holds its data.<\/p>\n
That building occupies land, draws power from an electricity grid and has to dissipate the heat produced by its machines. Technicians work there, supplies arrive and components are replaced. The service we use without a second thought also depends on these activities, which closely resemble the maintenance of any other industrial infrastructure.<\/p>\n
The word \u201ccloud\u201d has helped us forget all this. For someone using an application, that is convenient: nobody wants to learn about the server room before opening a document. For those choosing where to run a service, however, the location of the machines can affect performance, costs and the ability to recover after a failure.<\/p>\n
Asking where an online document is located can lead to more than one answer. Depending on the service, there may be a primary copy, replicas, backups and temporary versions. The software that processes it and the system that checks login credentials may also run on different infrastructures.<\/p>\n
Major providers organise this geography into clearly defined areas. AWS, for example, distinguishes between geographic Regions and the Availability Zones within them. A zone comprises one or more data centres, with redundant power and connectivity infrastructure. The chosen Region is therefore a useful reference point, but it does not correspond to the address of a single building.<\/p>\n
For a business, understanding this map means knowing where the different parts of the service reside. Finding the answer requires more care than simply selecting \u201cEurope\u201d from a menu.<\/p>\n
Between a phone and a data centre lies a network of connections. Some run beneath streets; others cross oceans. Submarine cables carry the vast majority of international data traffic: they are central to global connectivity, even though they feature far less prominently than satellites in our mental picture of the digital world.<\/p>\n
This network has specific landing points and routes. Two places that appear close on a map may be connected by a less direct route than expected. Geographic distance matters, but so does the way networks interconnect.<\/p>\n
The physical route helps explain latency: the time it takes for a request to receive a response. When reading an article, a slight additional delay may go unnoticed. During a video call, or in an application that makes many consecutive requests, the wait becomes more apparent. The speed of light remains a limit even when the contract promises an extremely fast service.<\/p>\n
A data centre needs favourable conditions around it. Available power, reliable connections and cooling options all help determine where it makes sense to build one.<\/p>\n
According to the International Energy Agency, cooling accounts for a variable share of data centres\u2019 electricity consumption, depending on the characteristics and efficiency of their systems. Water use also varies: some solutions consume water to dissipate heat, while others involve different trade-offs between water and energy use.<\/p>\n
Describing a service as digital therefore says little about its impact on the surrounding area. Assessing that impact requires information about the facility, the electricity it uses and local conditions. The same amount of water consumption carries different implications in an area with abundant supplies and one already experiencing water stress.<\/p>\n
Keeping an application close to its users can shorten part of the network journey. Concentrating all resources in one place, however, exposes them to the same local events.<\/p>\n
Distributing resources across separate zones helps contain this risk. This must be built into the architecture: the provider\u2019s availability of multiple sites does not automatically move an application between them or guarantee that it will keep running after a disruption. Replicas, connections and recovery procedures must be designed.<\/p>\n
A backup also deserves a geographical question. If it shares with the primary system the dependencies you are trying to protect against, it could become unreachable precisely when it is needed. Having a copy and being able to restore operations are different outcomes.<\/p>\n
The geography of the cloud should be considered alongside price when choosing a service. Where is the data stored? Where are the backups kept? Which dependencies remain if a site becomes unavailable? These are practical questions that need to be considered in relation to the product\u2019s features and configuration.<\/p>\n
Someone saving a photograph can continue to ignore this journey. Anyone putting a business management system, a learning platform or a public service online needs to understand at least its critical stages.<\/p>\n
The cloud allows us to use distant infrastructure without having to manage all of it directly. That distance remains, with its advantages and constraints. When \u201cSave complete\u201d appears on the screen, somewhere a machine has finished writing data. And it still needs electricity.<\/p>\n","protected":false},"excerpt":{"rendered":"
Server<\/p>\n","protected":false},"author":8,"featured_media":13659,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[39],"tags":[],"class_list":["post-13658","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-39"],"translation":{"provider":"WPGlobus","version":"3.0.0","language":"en","enabled_languages":["it","en"],"languages":{"it":{"title":true,"content":true,"excerpt":true},"en":{"title":true,"content":true,"excerpt":true}}},"yoast_head":"\n