bmwhonda Posted April 14, 2010 Posted April 14, 2010 I am running OSCommerce 2.2 now and will have about 300,000 items with 2000+ brand name, 1500 categories, 12,000 subcategories so the main product administration and search is painfully slow from the back end. Navigating to the subcategories to add/ edit product takes forever. It's running on 2- servers with 2 dual- Xeon (1 web/ OSCommerce and 1 database/ mySQL) so it's not the hardware. Is there a way to clean up some of the datafield or modify OSCommerce so it'll retrieve data faster? Everything is working fine but this slowness is making me considering another custom design site. Consulting service is welcome so PM me at bmwhonda @ hotmail.com. Thanks.
♥toyicebear Posted April 14, 2010 Posted April 14, 2010 Read this... Basics for osC 2.2 Design - Basics for Design V2.3+ - Seo & Sef Url's - Meta Tags for Your osC Shop - Steps to prevent Fraud... - MS3 and Team News... - SEO, Meta Tags, SEF Urls and osCommerce - Commercial Support Inquiries - OSC 2.3+ How To To see what more i can do for you check out my profile [click here]
MrPhil Posted April 14, 2010 Posted April 14, 2010 Is it only during Admin that it's painfully slow, or is it too slow for customers too? Either way, keep in mind that osC really wasn't designed to handle such huge amounts of data at once. A store built for 300,000 products would probably be architected a bit differently. You might look at the database -- perhaps using the InnoDB engine rather than the default MyISAM could speed you up. You might also look at something like EasyPopulate, to bulk read updates from a CSV file. It might not be any faster overall, but at least you could go off and do something else while waiting for the updates to be handled batch-style. If customers are experiencing slow performance too, InnoDB might be helpful there. If it doesn't help that much, only a hardware upgrade and/or a major rewrite of the software will save you. Maybe there's a consultant out there with lots of database experience who could point out redundant queries or badly structured queries that could be sped up. Perhaps queries in one place could be saved (cached) for reuse elsewhere, rather than being repeated (trade off space for time). Even the use of an interpreted scripting language (PHP) on something handling so much data might be suspect -- maybe you could look at moving some of the inner PHP code (doing a lot of data crunching) into compiled C extensions to PHP?
Recommended Posts
Archived
This topic is now archived and is closed to further replies.