raboof/pekko
1
0
Fork 0
Code Issues Pull requests Projects Releases 6 Packages Wiki Activity Actions 10

Merge pull request #1586 from akka/wip-2816-cluster-dispatcher-patriknw

Browse source 
DOC: Describe cluster dispatcher, see #2816
This commit is contained in:
Roland Kuhn 2013-07-04 06:46:02 -07:00
commit ad742feb00
2 changed files with 50 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

25
akka-docs/rst/java/cluster-usage.rst
View file

@ -376,6 +376,9 @@ Death watch uses the cluster failure detector for nodes in the cluster, i.e. it
generates ``Terminated`` message from network failures and JVM crashes, in addition generates ``Terminated`` message from network failures and JVM crashes, in addition
to graceful termination of watched actor. to graceful termination of watched actor.
If you encounter suspicious false positives when the system is under load you should
define a separate dispatcher for the cluster actors as described in :ref:`cluster_dispatcher_java`.
Cluster Aware Routers Cluster Aware Routers
^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^^^^
@ -739,3 +742,25 @@ Cluster Info Logging
You can silence the logging of cluster events at info level with configuration property:: You can silence the logging of cluster events at info level with configuration property::
akka.cluster.log-info = off akka.cluster.log-info = off
.. _cluster_dispatcher_java:
Cluster Dispatcher
------------------
Under the hood the cluster extension is implemented with actors and it can be necessary
to create a bulkhead for those actors to avoid disturbance from other actors. Especially
the heartbeating actors that is used for failure detection can generate false positives
if they are not given a chance to run at regular intervals.
For this purpose you can define a separate dispatcher to be used for the cluster actors::
akka.cluster.use-dispatcher = cluster-dispatcher
cluster-dispatcher {
type = "Dispatcher"
executor = "fork-join-executor"
fork-join-executor {
parallelism-min = 2
parallelism-max = 4
}
}

25
akka-docs/rst/scala/cluster-usage.rst
View file

@ -365,6 +365,9 @@ Death watch uses the cluster failure detector for nodes in the cluster, i.e. it
generates ``Terminated`` message from network failures and JVM crashes, in addition generates ``Terminated`` message from network failures and JVM crashes, in addition
to graceful termination of watched actor. to graceful termination of watched actor.
If you encounter suspicious false positives when the system is under load you should
define a separate dispatcher for the cluster actors as described in :ref:`cluster_dispatcher_scala`.
.. _cluster_aware_routers_scala: .. _cluster_aware_routers_scala:
Cluster Aware Routers Cluster Aware Routers
@ -761,3 +764,25 @@ Cluster Info Logging
You can silence the logging of cluster events at info level with configuration property:: You can silence the logging of cluster events at info level with configuration property::
akka.cluster.log-info = off akka.cluster.log-info = off
.. _cluster_dispatcher_scala:
Cluster Dispatcher
------------------
Under the hood the cluster extension is implemented with actors and it can be necessary
to create a bulkhead for those actors to avoid disturbance from other actors. Especially
the heartbeating actors that is used for failure detection can generate false positives
if they are not given a chance to run at regular intervals.
For this purpose you can define a separate dispatcher to be used for the cluster actors::
akka.cluster.use-dispatcher = cluster-dispatcher
cluster-dispatcher {
type = "Dispatcher"
executor = "fork-join-executor"
fork-join-executor {
parallelism-min = 2
parallelism-max = 4
}
}