Facebook架构和数据库ID以数万亿开头[关闭]

I have noticed that facebook has moved away from having id's that start with 1... there userIds and other objects like images etc start in the 100 trillions... Is there a reason they do this rather than starting at id 1? This is not exactly how fb are doing it, but the idea... users_id - start - 100'000'000'000 image_id - start - 200'000'000'000 video_id - start - 300'000'000'000

Is there a reason to do something like this?... I can see that this way every object would have a unique id, as even facebook would never get more than 100 trillion images and therefore the image_id would never cross over into the video_id, but why do it like that?

Most likely they do that for having the very simple way of identifying a particular object by only a single number. This also could help to shard the objects better.

But personally I've never seen any official explanation (though I read their blogs and see conferences)