Yeah, We Know...
Technically, you're not supposed to execute logic in threads within a JEE application. There are (will be?) JEE mechanisms available for threaded processing.
But sometimes, in some CDI applications, it makes sense to manage your own threads. Weft is an API that organizes and injects Executor instances for your CDI application.
One Annotation To Create Them All
Controlling which executor gets injected is simple, using org.commonjava.cdi.util.weft.ExecutorConfig:
@ExecutorConfig( named="my-threads" )
@Inject
private Executor executor;
In case your executor isn't configured (see Configuring, below), you can also set some default parameters in the annotation:
@ExecutorConfig( named="my-threads", daemon=true, priority=8, threads=10 )
[...]
Configuring
The ExecutorProvider component (which manages the Executor instances injected by Weft) @Inject's an instance of WeftConfig, which can specify priorities, thread counts, and the daemon flag for each named Executor your application uses. It also accepts a default priority and thread count setting to be applied to any Executor not otherwise configured. Settings in the injected WeftConfig instance always override those given in the @ExecutorConfig annotation, allowing the user to fine tune threading in the installed application.