Vitali Diatchkov ha scritto:
> 
> Just a hint from my mind related to our projects.
> 
> There are situations when 20 layers from Oracle are being rendered. The data
> is spread among the whole territory of Finland and the often situation is
> when there is no data to be rendered at the specified scale ratio for 17 of
> 20 layers. But every renderer not only requests features but also calls
> getCount() method to perform real checking to be sure that there is no data
> to be rendered if empty feature collection has come in the beginning.
> getCount() leads to "SELECT count(*) FROM.. WHERE.." and so on in case of
> OracleDataStore - for these 17 layers from 20. Totally we have 3 + (17 * 2)
> requests to the database (if we have 20 layers and only 3 of them contain
> features inside of current Filter). 

Eeek, getCount() on a big table can take tens of seconds (if you have 
millions of records).

> Did not investigate deeply but there
> should be a preference to switch off this getCount() call from renderer in
> case no features were loaded during usual request. 
> Isn't it the GT renderer calls this stuff, not UDIG wrappers ?... do not
> remember exactly, but anyway - it would be good optimization to switch off
> this call.

I could not find any getCount call in StreamingRenderer. Then I put a 
breakpoint in JDBCFEatureSource.count(xxx,xxx) and issued a GetMap 
request in GeoServer and the breakpoint was never reached, so I believe
there is nothing in StreamingRenderer asking for row count (all counting
code I know of in the end delegates to FeatoureSource.getCount(xxx)).
I would suggest you do the same with uDig and your datastore and see 
what code triggers this call. I'm curious, me too.

Cheers
Andrea

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
Geotools-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel

Reply via email to