Popular Get Request

As there is only four copies of any piece of data, what happens when a popular piece of data is requested by thousands/millions of users at the same time?

There is probably a simple answer.

1 Like

Every Vault along the X/OR route caches the chunk so the subsequent requests can be filled by vaults up-line rather than having to go all the way back to the source…

The more a file is requested, the more cached copies exist, and the faster the network performs.


Where does the cached data get stored? RAM or HDD?


Level 1 cache → RAM and then hdd. The cache levels will be dynamic but not initially (unless we can get it squeezed in). The idea is most recent cache is in RAM and then we have opportunities for very fast data access t recent data and in some cases I believe faster than local hdd (mg drive where transfer rates are 30-40 Mb/s). In code right now that cache I believe is on disk, but likely to be mmap file (issues with 32 bit though, of course).