Thu, 09 Apr 2026 12:39:02 +0000 https://bugs.horde.org/ticket/6403 Halve Kolab's Freebusy memory usage Hello together, attached patch almost halves Kolab's freebusy memory usage. 10.000 events now use 25mb instead of 44mb. I also experimented with splitting "$objects = store->getObjects()" into "$ids = store->getObjectIds(); foreach($object = store->getObjectById())" but it turned out to be slower and PHP5 seems to implement copy-on-write as it didn't make a difference from a memory usage point of view. I did some whitespace cleanup while I was at it... Cheers, Thomas PS: I almost queued this one for the framework package again ;-) Sat, 08 Mar 2008 09:28:34 +0000 https://bugs.horde.org/ticket/6403#t43571 This could probably be further optimized not using a foreach loop at all: http://www.horde.org/horde/docs/?f=CODING_STANDARDS.html#loops Sat, 08 Mar 2008 12:13:19 +0000 https://bugs.horde.org/ticket/6403#t43572 > Hello together, > > attached patch almost halves Kolab's freebusy memory usage. > 10.000 events now use 25mb instead of 44mb. Makes sense. Thanks! In CVS. I didn't trace this stuff before so now that I looked at it I'd like to suggest the attached patch in addition to your fix. Can you check it on your side? There is probably more potential to get the thing less memory/cpu time hungry. > > I also experimented with splitting "$objects = store->getObjects()" > into "$ids = store->getObjectIds(); foreach($object = > store->getObjectById())" > but it turned out to be slower and PHP5 seems to implement > copy-on-write as it didn't make a difference from a memory usage > point of view. > > I did some whitespace cleanup while I was at it... > > Cheers, > Thomas > > PS: I almost queued this one for the framework package again ;-) > Tue, 11 Mar 2008 10:23:33 +0000 https://bugs.horde.org/ticket/6403#t43699 > I didn't trace this stuff before so now that I looked at it I'd like > to suggest the attached patch in addition to your fix. Can you check > it on your side? Wow, I first thought your change wouldn't make much difference but now it takes only 0.983s instead of 2.340s for 10.000 events on a P4 3 GHz :-) Yeah. Tue, 11 Mar 2008 15:32:23 +0000 https://bugs.horde.org/ticket/6403#t43709 >> I didn't trace this stuff before so now that I looked at it I'd like >> to suggest the attached patch in addition to your fix. Can you check >> it on your side? > > Wow, I first thought your change wouldn't make much difference but > now it takes only 0.983s instead of 2.340s for 10.000 events on a P4 > 3 GHz :-) Yeah. > Cool, thanks for testing. Closing this one then. Tue, 11 Mar 2008 17:04:56 +0000 https://bugs.horde.org/ticket/6403#t43714 in cvs Tue, 11 Mar 2008 17:05:09 +0000 https://bugs.horde.org/ticket/6403#t43715