Loading repository data…
Loading repository data…
Attriumph / repository
:fist: A full-stack web application, developed by MEAN Stack, that generates short URL and shows the statistics data of the usage of the short URL; Enhanced independence by deploying Docker and improved QPS by using Redis as cache and Nginx as reverse proxy.
We assume daily active user： 1,000,000
(1) average estimate
for Insert：
for Lookup：
(2) peek estimate:
Therefore, we can deal with it by a single machine
map<longURL, shortURL> LongToShortMap;
map<shortURL, longURL> ShortToLongMap;
shortURL insert(longURL){
if LongToShortMap not containsKey longURL:
generate shortURL;
put<longURL, shortURL> into LongToShortMap;
put<shortURL, longURL> into ShortToLongMap;
return LongToShortMap(longURL);
}
longURL lookup(shortURL){
return ShortToLongMap.get(shortURL);
}
The simplest way: return map.size()，but the size will increase too big, so we need add letters. Therefore, we need to convert decimal into 62 hex. Here is the code:
shortURL generateShortURL(){
return convertTo62(ShortToLong.size());
}
shortURL convertTo62(mapSize){
char encode[62] = {'a',...'z','A',...'Z','0',...'9'};
shortURL = "";
while(mapSize != 0){
shortURL = encode[mapSize % 62] + shortURL;
mapSize /= 62;
}
return shortURL;
}
According to experience, we can assume:
Average size of long url = 100 byte
Average size of short url = 4 bytes(int)
Daily new URL = 104 * 100,000 = 104,000,000 byte = 104mb
Year new URL = 104mb * 365 = 37960mb < 40GB
Therefore, singe Machine can handle!