The best solution I've found so far is defining a "poison" task like the
following one (Groovy code):
static class PoisonPill implements IgniteRunnable {
@IgniteInstanceResource
private transient Ignite ignite
@Override
void run() {
ignite.close()
}
}
and broadcasting it to all remote nodes:
def void shutdown() {
grid.cluster().forRemotes().ignite().compute().broadcast(new
PoisonPill())
}
That works but in the "master" I'm getting the following exception:
java.lang.IllegalStateException: Grid is in invalid state to perform this
operation. It either not started yet or has already being or have stopped
[gridName=nextflow, state=STOPPING]
at
org.apache.ignite.internal.GridKernalGatewayImpl.illegalState(GridKernalGatewayImpl.java:190)
~[ignite-core-1.4.0.jar:1.4.0]
at
org.apache.ignite.internal.GridKernalGatewayImpl.readLock(GridKernalGatewayImpl.java:90)
~[ignite-core-1.4.0.jar:1.4.0]
at
org.apache.ignite.internal.cluster.ClusterGroupAdapter.guard(ClusterGroupAdapter.java:170)
~[ignite-core-1.4.0.jar:1.4.0]
at
org.apache.ignite.internal.cluster.ClusterGroupAdapter.forPredicate(ClusterGroupAdapter.java:367)
~[ignite-core-1.4.0.jar:1.4.0]
at
org.apache.ignite.internal.cluster.ClusterGroupAdapter.forOthers(ClusterGroupAdapter.java:564)
~[ignite-core-1.4.0.jar:1.4.0]
at
org.apache.ignite.internal.cluster.ClusterGroupAdapter.forRemotes(ClusterGroupAdapter.java:536)
~[ignite-core-1.4.0.jar:1.4.0]
at org.apache.ignite.cluster.ClusterGroup$forRemotes$1.call(Unknown Source)
~[na:na]
at nextflow.executor.IgConnector.shutdown(IgConnector.groovy:155)
~[nxf-ignite-0.16.0-SNAPSHOT.jar:na]
Is there a better way to shut-down all cluster instances ?
Cheers,
Paolo
On Wed, Sep 30, 2015 at 4:03 PM, Paolo Di Tommaso <[email protected]
> wrote:
> Hi,
>
>
> Is there an API to gently shut-down all the nodes that made up an Ignite
> cluster?
>
>
> Cheers,
> Paolo
>
>
