You can trade time spent to load the hash into memory against time spent accessing an element by using a tied hash like DB_File, or, if your hash is basically constant, CDB_File. But that will slow down every access, so you will have to benchmark whether all your hash accesses will sum up to the 7 seconds load time. You could also investigate whether simply storing the hash as Perl code via Data::Dumper and loading it via do will be faster (I doubt it), or whether using/writing a daemon like Memcached improves things.
In reply to Re: storable: too slow retrieval of big file.
by Corion
in thread storable: too slow retrieval of big file.
by pc2
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |