When a frontend is deployed, there are various assets such as images, videos, svg’s, etc that are used but those assets cannot be fetched directly from a mongo or postgres database.
For such assets like images, videos, etc, it is a good practice to create object stores. These are databases for multimodal data.
Aws provides such object stores called as S3 as a service → Simple Storage Service.
The method to create an object store in aws via s3 is by creating buckets. Per say you have multiple clients → NComputing, Manasvi Foundation, Google, etc using your service. You create a bucket for them each to store their objects and use a public-private key to access the content of the bucket.
These networks are used to efficiently manage the load put on a server where the website is hosted. It does it by creating copies of the websites at nearest servers and the local clients access the website from there, let me explain by an example.
Consider that you have made a website www.kahaani.site. Now if there are 50 people in Australia, 50 in India and 50 in the US → do you think if they all hit at your server which is present in India be able to serve their requests in the same latency ? The answer is no → The clients sitting in India would have low latency while those in the US and Australia will have high latency, this is because distance is directly proportional to latency, now how do we solve this?
We solve this by using a multi-server architecture, when a user in the US reqs data from kahaani.site in India, a replica of the website is cached in the US server for some however and now whenever another user hits the same requests, the response is given by the US Server hosting the website which reduces the latency and this architecture can be auto-scaled and if the same happens in other countries is when the same algorithm will be used to reduce the latency and if nobody hits the website in a specific region, the cache is cleared.
To sum if up, the benefits are:
CDN Providers I like: