patrik wrote:
Odd, I'm running 0.8 on FC5, and only noticed this last night. The even
odder thing in my case was it only applied to parse plugins. A urlfilter
only specified in the nutch-site.xml was fine.
What you all describe is quite unlikely ... Config resources are loaded
in a specific order, and if this modified nutch-site.xml was read at all
the properties specified there will always override values specified in
nutch-default.xml.
The only possible explanations that come to my mind are these:
* you misspelled a property name in nutch-site.xml - it may happen.
* you changed the file in WEB-INF/classes, and then reloaded the webapp,
and your servlet container re-deployed the app from the original war
file, thus overwriting your modified file.
* you have other copies of (unchanged) nutch-site.xml on your classpath,
or inside jars loaded on your classpath, with identical name (basically,
"./nutch-site.xml", which take precedence over your modified nutch-site.xml
You can change your log4j.properties to activate DEBUG level (or use
-Dhadoop.root.logger=DEBUG,console -Dnutch.root.logger=DEBUG on the
command-line), and check the log - there should be an information which
config files are loaded in what order and from what locations.
--
Best regards,
Andrzej Bialecki <><
___. ___ ___ ___ _ _ __________________________________
[__ || __|__/|__||\/| Information Retrieval, Semantic Web
___|||__|| \| || | Embedded Unix, System Integration
http://www.sigram.com Contact: info at sigram dot com
