Agree, I think we should make optimizations before deployment not after. Of course some performance tests can't be done on dev environment, but those tests shouldn't use too much CPU. Most hosting companies doesn't have tremendous servers like google.
On database driven systems (i.e., Drupal, Wordpress), native optimizations seem to do better that mod_pagespeed. A test Wordpress site was about 50% faster using W3 Total Cache than with mod_pagespeed. However, a static php/html site on the same server was about 70% faster.
I started playing with mod_pagespeed on my new server and very interesting indeed for php/mysql based site. I get 2x times faster page loading speed but at the cost of 3x times the cpu load and that is with image rewrites disabled already. Interested in finding where the cpu load comes from.
Please stress that mod_pagespeed is not for all sites, I tried it on mine and it slowed the site down and drastically increased the server load.
ReplyDeleteAgree, I think we should make optimizations before deployment not after.
DeleteOf course some performance tests can't be done on dev environment, but those tests shouldn't use too much CPU. Most hosting companies doesn't have tremendous servers like google.
worthy info
ReplyDeletethanks
ya useful i should say.
ReplyDeleteOn database driven systems (i.e., Drupal, Wordpress), native optimizations seem to do better that mod_pagespeed. A test Wordpress site was about 50% faster using W3 Total Cache than with mod_pagespeed. However, a static php/html site on the same server was about 70% faster.
ReplyDeleteI started playing with mod_pagespeed on my new server and very interesting indeed for php/mysql based site. I get 2x times faster page loading speed but at the cost of 3x times the cpu load and that is with image rewrites disabled already. Interested in finding where the cpu load comes from.
ReplyDelete