Up to Karaf 3, the Karaf JDBC feature created a datasource service using blueprint (as you do).
It means that: 1. The JDBC feature had a dependency to blueprint 2. Changing the datasource configuration itself triggered unregister/register of the datasource service (as the blueprint container is "reloaded"). That's why we created pax-jdbc and Karaf 4.x now uses pax-jdbc: 1. pax-jdbc uses ConfigAdmin/cfg to describe the datasource configuration, and use a managed factory to create the corresponding datasource service: we don't have the blueprint dependency anymore. 2. the managed factory allows to reload only the datasource, not necessary the service itself. The usage of the datasource service itself don't change (it's a service that you can use with blueprint, ds, or pure OSGi). Exposing a datasource service using blueprint is still supported but you have to install the aries-blueprint feature. Regards JB On 11/06/2018 18:35, ceugster wrote: > What do you mean by > jbonofre wrote >> By the way, why don't you use pax-jdbc ? We removed the blueprint >> approach for pax-jdbc to avoid the blueprint dependency and provide a >> more dynamic approach. > > ? Are there bundles that I have to uninstall and install? > > Regards > Christian > > > > > -- > Sent from: http://karaf.922171.n3.nabble.com/Karaf-User-f930749.html > -- Jean-Baptiste Onofré [email protected] http://blog.nanthrax.net Talend - http://www.talend.com
