Then, let's say there is a million refs, each transaction involves only a couple of refs, and there is a number of concurrent transactions typical for a web application that runs on one server (let's say 10-100). Of course that the real answer is in measuring it, but I wanted to know, before I build it, whether such design is feasible from the standpoint of the refs design.
On Dec 11, 1:19 am, Phil Hagelberg <p...@hagelb.org> wrote: > DraganDjuric <draga...@gmail.com> writes: > > Is it a good idea, from the performance standpoint, to use many refs > > in the program? By many, I mean thousands or even millions? > > The question isn't how many refs you should instantiate, it's how many > refs you should read or write to at once in a transaction as well as how > many transactions you expect to run at once. The place you'll run into > perf issues is with transaction retries. The best (only?) way to know is > to measure, I think. > > -Phil -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
