[TYPO3-dev] performace of cachingframework

Peter Russ peter.russ at 4many.net
Fri Feb 18 18:32:17 CET 2011


--- Original Nachricht ---
Absender:   Bjoern Pedersen
Datum:       18.02.2011 17:15:
> Am 18.02.2011 11:51, schrieb Peter Russ:
>> --- Original Nachricht ---
>>
>> This assumption is correct on your machine. But on high load sites where
>>  values are cached for some secs or mins it looks totally different:
>> There could be several 100s or 1000s of records with the same key before
>> GB....
>>
>> It would help a lot if developers would realize that TYPO3 claims to be
>> an "Enterpise CMS" and this more than just "runs fine on my machine" ;-)
>>
>> Peter.
>>
>>
> Yes, that is valid point.
> 
> A few more infos that I need from the cache specialists:
> is the has-function actually used anywhere except in some BE's
> internally? A grep on the core sources did not find anything. There were
> just hits in the tests and in swiftmailer (implementing it's own cache)
> and in extbase, also not using the normal cache.
> 
> 
> If it is not really used, i would suggest to deprecate it to avoid a
> usage pattern like:
>  if(cache->has(id)) then cache->get(id).
> 
> Björn

:-)
1#: Caching Framework (CFW) is a backport from the "famous" FLOW3 
foam-work*. Due to the fact that it has a nice architecture at the cost 
of performance this implementation of the CFW come up.

2nd: If you want to change something in the CFW you should talk to the 
MoDs* of FLOW3. To get some changes into CORE it takes 6 month to 
several years. No idea about FLOW3 with it's academical approach*.

Enjoy your weekend.

Peter.


*Just my personal opinion. To get me right: FLOW3 has an excellent 
design/architecture. But it's slow by design. But may be I'm wrong ;-)

-- 
docendo discimus

_____________________________
uon GbR
http://www.uon.li
http://www.xing.com/profile/Peter_Russ




More information about the TYPO3-dev mailing list