No responses? Should I update the wiki, create issues, and submit patches?
Tim Moloney The reasonable man adapts himself to MRSL the world; the unreasonable one persists 2015 Cattlemen Road in trying to adapt the world to himself. Sarasota, FL 34232 Therefore all progress depends on the (941) 377-6775 x208 unreasonable man. George Bernard Shaw > -----Original Message----- > From: Moloney, Tim M [mailto:[email protected]] > Sent: Saturday, June 06, 2009 18:59 > To: [email protected] > Subject: RE: Configuring a ManagedServiceFactory from a feature > > > I did misunderstand something. > > When using a ManagedServiceFactory the value for the name attribute is > "servicePid-instanceId", where servicePid is the service pid of the > ManagedServiceFactory and instanceId is a label used to uniquely > identify a particular service instance. > > If this is correct, I can update the Karaf Provisioning page > (http://felix.apache.org/site/46-provisioning.html). > > There also appears to be a bug in FeaturesServiceImpl.java. I think > that line 439 should be changed from > > ... ALIAS_KEY + "=" + pid ... > > to > > ... ALIAS_KEY + "=" + factoryPid ... > > so that existing configurations can be retrieved. > > Several follow on comments/questions. > > - I was originally confused by the variable named > "factoryPid". A less > confusing name would be instanceId since it is definitely not the > factory service pid. > > - The filter used to get the list of configurations searches for a > property name ".alias_factory_pid". However checkKey() in the > ConfigurationAdmin's CaseInsensitiveDictionary only allows > keys to start > with an alpha-numeric character and has an addition test to > prevent keys > from starting or ending with a period. Therefore the filter can be > simplified from > > (|(ALIAS_KEY=factoryPid)(.alias_factory_pid=factoryPid)) > > to just > > (ALIAS_KEY=factoryPid) > > BTW, this same filter also appears in FileMonitor.java in Karaf's > deployer. > > - Would it be reasonable to add the "instanceId" (as I called > it above) > to the displayed bundle properties in the webconsole? > > - Should the feature's configurations be deleted when it is > uninstalled? > > If everything that I said above is correct/reasonable, I'll > create JIRA > issues and start working on patches. > > > Tim Moloney The reasonable man adapts himself to > MRSL the world; the unreasonable one persists > 2015 Cattlemen Road in trying to adapt the world to himself. > Sarasota, FL 34232 Therefore all progress depends on the > (941) 377-6775 x208 unreasonable man. George Bernard Shaw > > > > > -----Original Message----- > > From: Moloney, Tim M [mailto:[email protected]] > > Sent: Saturday, June 06, 2009 14:19 > > To: [email protected] > > Subject: Configuring a ManagedServiceFactory from a feature > > > > > > I was figuring out why my ManagedServiceFactory wasn't getting > > configured by the feature when I discovered the odd notation used to > > specify the service pid. > > > > I found that if the name attribute of the config element > has a dash in > > it, the attribute value will be split into a service pid > and a service > > factory pid. If I read the code correctly, both the service > > pid and the > > factory service pid need to be set to the name of the > > ManagedServiceFactory. So I managed to get my ManagedServiceFactory > > configured by specifying: > > > > <config > > name="com.foo.MyManagedServiceFactory-com.foo.MyManagedService > > Factory"> > > ... </config> > > > > However that seems a bit verbose and is just one more place > > where a typo > > can stop things from working. I think it would be better if > > there was a > > different attribute name to configure ManagedServiceFactories. For > > example, configuring a ManagedService would stay the same: > > > > <config name="com.foo.MyManagedService"> ... </config> > > > > But configuring a ManagedServiceFactory would be: > > > > <config factoryName="com.foo.MyManagedServiceFactory"> > ... </config> > > > > Is this a good idea or do I misunderstand something? > > > > > > Tim Moloney The reasonable man adapts himself to > > MRSL the world; the unreasonable one persists > > 2015 Cattlemen Road in trying to adapt the world to himself. > > Sarasota, FL 34232 Therefore all progress depends on the > > (941) 377-6775 x208 unreasonable man. George Bernard Shaw > > > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
