[TYPO3-dev] Typo3 Performance

Steffen Kamper steffen at dislabs.de
Fri Mar 2 13:44:12 CET 2007


"R. van Twisk" <typo3 at rvt.dds.nl> schrieb im Newsbeitrag 
news:mailman.48403.1172798616.21067.typo3-dev at lists.netfielders.de...
> Dmitry Dulepov wrote:
>> Martin Kutschker wrote:
>>
>>> PS: Best improvement of course is to reduce the total amount of queries 
>>> used :-)
>>>
>>
>> Big +1. Why don't we have generic query cache in typo3? Some queries 
>> repeat in BE/FE a number of times. Btw, I like your rootline cache idea 
>> in core list but I think that db cache should be more on t3lib_db 
>> level...
>>
>>

this remembers me to my question about implementing nested sets in pages 
table (e.g.).
Instead of doing recursiv function and sql-query each time it could be 
solved with one query instead.

May be we should look for improving the queries and tabledesign of the core, 
i think there is enough potential to improve a lot of speed. In my opinion 
this is the more efficient way instead of "dirty queries" and afterword 
tuning up the system.

vg  Steffen

> For mysql databases I am not sure how much time it will really save,
> since mysql has it's own query cache already. This is the same for other
> databases of course which uses other methods to cache results.
>
> I have done some limited testing using adodb, DBAL
> and using CacheExecute and couldn't find a big improvement yet,
> at least not to justify the investigation time...
>
> I think we can get more speed out of setting indexes correctly, we might 
> consider
> a storing all queries in a database and the time it took to execute the 
> query
> or  filesystem (option can be turned off or on) and then based on the 
> acquired data
> we can create some statistics . Based on that it might shed some light
> on where to place indexes.
>
>
> Ries
>
> Note: My testing was done about a couple of months after I started to use
> typo3 about 2.5 years ago. The result may not be true anymore.
> 






More information about the TYPO3-dev mailing list