Isn't that a real problem if the factory is already set ? Or is that a valid use case ?
On Fri, Mar 2, 2012 at 10:52, David Bosschaert <[email protected]>wrote: > Another email about logging ... > > During the integration of the Aries JNDI component in our project we > hit an issue with the log messages emitted from > > https://svn.apache.org/repos/asf/aries/trunk/jndi/jndi-core/src/main/java/org/apache/aries/jndi/startup/Activator.java > > Specifically the following block (there is a second block just like > this one too): > > try { > OSGiInitialContextFactoryBuilder builder = new > OSGiInitialContextFactoryBuilder(); > NamingManager.setInitialContextFactoryBuilder(builder); > icfBuilder = builder; > } catch (NamingException e) { > LOGGER.info(Utils.MESSAGES.getMessage("unable.to.set.static.ICFB"), e); > } catch (IllegalStateException e) { > > > LOGGER.info(Utils.MESSAGES.getMessage("unable.to.set.static.ICFB.already.exists", > getClassName(InitialContextFactoryBuilder.class)), e); > } > > In our case the InitialContextFactoryBuilder is already set, so there > will be an IllegalStateException, but this is totally harmless in our > case. It's fine to log an INFO message, but we don't want the > exception to appear at this level, as that gets people worried, > therefore, would it be acceptable to change the log message to > something like this? > > > > LOGGER.info(Utils.MESSAGES.getMessage("unable.to.set.static.ICFB.already.exists", > getClassName(InitialContextFactoryBuilder.class))); > > > LOGGER.debug(Utils.MESSAGES.getMessage("unable.to.set.static.ICFB.already.exists", > getClassName(InitialContextFactoryBuilder.class)), e); > > So that the message is still logged at debug level but the exception > only appear at Debug level? > > Thanks, > > David > -- ------------------------ Guillaume Nodet ------------------------ Blog: http://gnodet.blogspot.com/ ------------------------ FuseSource, Integration everywhere http://fusesource.com
