Focusing on read performance , I want to create a term, such as Orddict or Proplist, that contains a large number of entries (100,000s), each of which contains an identifier and a Term value. This encapsulating term should be able to return the value stored under its key, just as Orddik can do.
example:
K001 - Term001
K002 - Term002
K003 - Term003
The resulting Term containing the entire set must be transferred from function to function for several computational purposes, without storing it in a persistence store in order to avoid disk I / O. I also decided not to use memory caching at this point in order to avoid the architectural complexity at the moment, so I am focused on ensuring that all this is simply key to the search.
Orddicts are sorted by keywords, which increases key search compared to regular dict. I do not know of any other Erlang module that could implement a more efficient indexing mechanism in its Term.
Any suggestions for an approach better than a decree?
source
share