[TYPO3-dev] Typo3 Database design question (optimization / indices)

Marc Wöhlken woehlken at quadracom.de
Thu Mar 23 20:35:53 CET 2006


Ernesto Baschny [cron IT] schrieb:
> Do you have any statistics on what percentage of processing time is
> spend in the SQL-queries? If we are talking about <2%, then I would
> understand why noone has yet tried improving *that* part, since there
> are probably other "chunks" that are better candidates for optimizing.

As I just did some simple tests yesterday I have not that kind of data.
This is what I want to analyze in my next step.


>> Than I remeasured the db processing time and found it to be remarkably
>> faster: 0.0004s on average!
> 
> In my tests, I see no difference in the "parse-time" (shown by TYPO3 at
> the bottom of the page, if debug is on) when calling some typo3-page
> with and without those indexes with "&no_cache=1".
> 
> But in your case you should see a reduction of 122ms in parse-time, right?

Let's make one thing clear: This is just initial testing. It's not the
optimal way to do test but rather a first quick and easy (aka dirty)
approach ;-)

When I measured db performance I took the db processing of some cached
pages as cached pages seem more natural to me (most ppl use caching).

To find db processing times I took the pagewise fastest measured
processing time of several pages. To make it clear: I measured the
processing time of one page about 100 times. The fastest time was recorded.
I repeated this process with indexes added.

This is how 0.122s vs. 0.0004s came to my mind.

Hope some things cleared up

Regards
 Marc





More information about the TYPO3-dev mailing list