Hello Simone,

On 18-03-15 15:23, Simone Giannecchini wrote:
> I will ask Nicola to clarify further. This work was performed in order 
> to make GeoServer work nicely when JDBCConfig is used and there is a 
> ton of layers around (order of 100K). This means, if you make a fix we 
> need to make sure the fix does not jeopardize performance. 

The fix doesn't influence performance. I didn't change anything about 
the fact that filters are being used for performance. The filter is just 
being wrongly created at the moment, it includes a filter that shouldn't 
be there. If anything my change improves performance because there is 
sometimes one filter less.

> I will ask Nicola to coordinate also in order to do some live testing 
> with an instance where we have thousands of layers. As I said this 
> fix, despite to the problem you found is crucial for performance 
> reasons in some (real world use cases).

That's cool, but I think in the first place there should be a simple 
test (it does not need many layers) that merely confirms that the 
conversion from tree to filter is being done correctly. At the moment 
there is no such test. This new method has been added and became part of 
the security system of geoserver; when this very crucial change was made 
a test should have been added that proves the correctness of this 
method, which unfortunately didn't happen.

Regards
Niels


------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the 
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Geoserver-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-devel

Reply via email to