Front page | perl.perl5.porters |
Postings from October 2009
Like others have suggested, if possible, taking advantage of your Pg database instead of using huge perl data structures would be a good approach. If your code can't accommodate that, what about mod_perl or fastcgi? On Tue, Oct 27, 2009 at 11:50 AM, Peter Keen <peter.keen@bugsplat.info>wrote: > Hi guys, > > I'm writing a web application that needs to thaw between 50 and 100 > frozen objects per web-request. They come in three types: > > - A simple hash of string to int with about 30k pairs > - A single string that is treated as a bitset using vec() that is > around 12k max. Most are smaller. > - A list-of-lists of about 30k elements, where each element is > composed of a string and two ints. > > Each of these is frozen, compressed with Gzip, and stuck in a > postgresql bytea column. Currently, half of the time spent on each web > request is spent thawing objects which is sort of frustrating. I'm > wondering if there's a way to make Storable faster, or if there is > something faster that I can use that is still perl. I'm about to throw > up my arms and embark on writing a C++ backend web service for this > but I figured I would ask you all first. > > Thanks, > Pete >Thread Previous | Thread Next