When we download applications or software from a server, why are we often encouraged to use a mirror site, rather than the “original”?
In: Technology
It’s a way to load balance server load amount multiple servers. Chances are if you are being pushed to a mirror it’s because it’s over loaded. CDN’s are a better way of handling this tbh.
Let’s say you owned a store that became really popular. The store only had room for 100 people in it at a time, but as your store becomes more and more popular, people start having to wait outside.
You try to tweak the inside of the store and maybe you squeeze a few more people in, but the only way to expand the current store further is to buy the place next door which is REALLY expensive ($1 million dollars). If you did that, you could handle 200 people but then once you hit that limit, you are faced with the same problem and expensive costs.
Plus, you have people driving all the way from other nearby cities to come to you, so their trip is lengthy just to buy from your store.
You realize that for the same cost ($1 million dollars), you could set up TWO additional stores in nearby cities. So instead of one store that can handle 200 people, you end up with three stores that each serve 100 people. And since the stores are closer to some of the shoppers, it is no longer a really long trip for them.
And if you ever need more stores, it is much easier, cheaper, and quicker to buy more stores instead of trying to expand the existing stores.
This is the same thing with servers and mirrors. Mirrors are the additional servers that provide the same files, and can provide them at faster download speeds if there is a mirror that is closer to wherever you are. And faster downloads also means the download is FINISHED faster, so that also helps create more availability on each mirror.
And on top of it all, it increases reliability because if one server has a critical failure, people can go to a mirror instead of getting mad about being unable to download the file.
When you download something, that means a server has to send that file from its own drive, through the internet to you.
So every download is taking up a certain amount of processing power and bandwidth that the server has access to.
If there are a lot of people wanting the same data at the same time (or wanting other data from the same server), that can mean the server hits its limits and has to ration the bandwidth given to each file – less bandwidth means it takes longer to download.
One solution to this is to setup a mirror – so instead of downloading from the main server, you can download it from a copy of that server that has its own processing power and bandwidth. Because you now have all the downloads being shared between multiple servers, there is less chance of hitting the limits of each one and having the performance limited.
Latest Answers