JSMin::get, JSMin::action, JSMin::text, and JSMin::min account for 43% of the exclusive wall time and 47% of exclusive cpu time.

I have been digging in using Xhprof to get these stats and have tracked it to javascript_aggregator.

Is there anything that can be done about the cpu % used on a page request?

And please educate me if I have read this out put incorrectly.

thanks

CommentFileSizeAuthor
xhprof-4.jpg100.76 KBsocialnicheguru
xhprof-3.jpg114.32 KBsocialnicheguru
xhprof-2.jpg109.35 KBsocialnicheguru
xhprof-1.jpg197.94 KBsocialnicheguru

Comments

danepowell’s picture

danepowell commented

Woah... I would certainly like an answer to this as well. I've never seen any performance issues related to Javascript Aggregator but I've never actually checked like this. This is a little disturbing.

mstrelan’s picture

mstrelan commented

I am getting very similar results, and have just disabled JSMin. Initial tests seem to be almost twice as fast.

philsward’s picture

philsward commented

I'll agree that this module creates a huge performance problem... I noticed it mainly on any logged in admin pages vs anonymous. Boost from the anonymous side of things seemed to pretty well take care of things, but when I was logged in, I was sitting 20-40 seconds to edit a node, or navigate the site. I thought maybe it had something to do with jsmin+ vs jsmin, but I think the results were pretty close to the same, regardless of which one was used.

jonaswouters’s picture

jonaswouters commented

Same experience here. There is definitely a performance problem with this module.