Use asynchronous clojureql or STM + entries?

Question

I am writing an application in clojure that should save my data, but if some data is lost, then this is not catastrophic. I am wondering if I should use ClojureQL every time I want to access data, or use STM + ClojureQL and transfer data to updates / records asynchronously. Does anyone know any advantages / disadvantages of each approach. And has anyone tried an asynchronous approach to conservation?

Answer 1

This is a classic approach to improving write latency, often called write-behind. Combining updates and executing them in a separate asynchronous stream has several advantages:

• ( ) .
• ,
• ( , , , )

:

• ,
• .
• , "" ; write-behind , .
• store-write , - . - write-behind , .

ClojureQL STM, .:)