Migration Guide 2.2.x to 2.3.x
The 2.3 release contains some structural changes that require some simple, mechanical source-level changes in client code.
When migrating from earlier versions you should first follow the instructions for migrating 1.3.x to 2.0.x and then 2.0.x to 2.1.x and then 2.1.x to 2.2.x.
Removed hand over data in cluster singleton
The support for passing data from previous singleton instance to new instance
in a graceful leaving scenario has been removed. Valuable state should be persisted
in durable storage instead, e.g. using akka-persistence. The constructor/props parameters
of ClusterSingletonManager
has been changed to ordinary Props
parameter for the
singleton actor instead of the factory parameter.
Changed cluster auto-down configuration
akka.cluster.auto-down
setting has been replaced by akka.cluster.auto-down-unreachable-after
,
which instructs the cluster to automatically mark unreachable nodes as DOWN after this
configured time of unreachability. This feature is disabled by default, as it also was in 2.2.x.
During the deprecation phase akka.cluster.auto-down=on
is interpreted at as instant auto-down.
Routers
The routers have been cleaned up and enhanced. The routing logic has been extracted to be usable within normal actors as well. Some usability problems have been have been solved, such as properly reject invalid configuration combinations. Routees can be dynamically added and removed by sending special management messages to the router.
The two types of routers have been named Pool
and Group
to make them more distinguishable and reduce confusion
of their subtle differences:
- Pool - The router creates routees as child actors and removes them from the router if they terminate.
- Group - The routee actors are created externally to the router and the router sends messages to the specified path using actor selection, without watching for termination.
Configuration of routers is compatible with 2.2.x, but the router
type should preferably be specified
with -pool
or -group
suffix.
Some classes used for programmatic definition of routers have been renamed, but the old classes remain as
deprecated. The compiler will guide you with deprecation warning. For example RoundRobinRouter
has
been renamed to RoundRobinPool
or RoundRobinGroup
depending on which type you are actually using.
There is no replacement for SmallestMailboxRouter
combined with routee paths, i.e. a group, because that
combination is not useful.
An optional API enhancement that makes the code read better is to use the props
method instead of withRouter
.
withRouter
has not been deprecated and you can continue to use that if you prefer that way of defining a router.
Example in Scala:
context.actorOf(FromConfig.props(Props[Worker]), "router1")
context.actorOf(RoundRobinPool(5).props(Props[Worker]), "router2")
Example in Java:
getContext().actorOf(FromConfig.getInstance().props(Props.create(Worker.class)),
"router1");
getContext().actorOf(new RoundRobinPool(5).props(Props.create(Worker.class)),
"router2");
To support multiple routee paths for a cluster aware router sending to paths the deployment configuration
property cluster.routees-path
has been changed to string list routees.paths
property.
The old cluster.routees-path
is deprecated, but still working during the deprecation phase.
Example:
/router4 {
router = round-robin-group
nr-of-instances = 10
routees.paths = ["/user/myserviceA", "/user/myserviceB"]
cluster.enabled = on
}
The API for creating custom routers and resizers have changed without keeping the old API as deprecated. That should be a an API used by only a few users and they should be able to migrate to the new API without much trouble.
Read more about the new routers in the documentation for Scala and documentation for Java.
Akka IO is no longer experimental
The core IO layer introduced in Akka 2.2 is now a fully supported module of Akka.
Experimental Pipelines IO abstraction has been removed
Pipelines in the form introduced by 2.2 has been found unintuitive and are therefore discontinued. A new more flexible and easier-to-use abstraction will replace their role in the future. Pipelines will be still available in the 2.2 series.
Changed cluster expected-response-after configuration
Configuration property akka.cluster.failure-detector.heartbeat-request.expected-response-after
has been renamed to akka.cluster.failure-detector.expected-response-after
.
Removed automatic retry feature from Remoting in favor of retry-gate
The retry-gate feature is now the only failure handling strategy in Remoting. This change means that when remoting detects faulty
connections it goes into a gated state where all buffered and subsequent remote messages are dropped until the configurable
time defined by the configuration key akka.remote.retry-gate-closed-for
elapses after the failure event. This
behavior prevents reconnect storms and unbounded buffer growth during network instabilities. After the configured
time elapses the gate is lifted and a new connection will be attempted when there are new remote messages to be
delivered.
In concert with this change all settings related to the old reconnect behavior (akka.remote.retry-window
and
akka.remote.maximum-retries-in-window
) were removed.
The timeout setting akka.remote.gate-invalid-addresses-for
that controlled the gate interval for certain failure
events is also removed and all gating intervals are now controlled by the akka.remote.retry-gate-closed-for
setting
instead.
Reduced default sensitivity settings for transport failure detector in Remoting
Since the most commonly used transport with Remoting is TCP, which provides proper connection termination events the failure detector sensitivity
setting akka.remote.transport-failure-detector.acceptable-heartbeat-pause
now defaults to 20 seconds to reduce load induced
false-positive failure detection events in remoting. In case a non-connection-oriented protocol is used it is recommended
to change this and the akka.remote.transport-failure-detector.heartbeat-interval
setting to a more sensitive value.
Quarantine is now permanent
The setting that controlled the length of quarantine akka.remote.quarantine-systems-for
has been removed. The only
setting available now is akka.remote.prune-quarantine-marker-after
which influences how long quarantine tombstones
are kept around to avoid long-term memory leaks. This new setting defaults to 5 days.
Remoting uses a dedicated dispatcher by default
The default value of akka.remote.use-dispatcher
has been changed to a dedicated dispatcher.
Dataflow is Deprecated
Akka dataflow is superseded by Scala Async.
Durable Mailboxes are Deprecated
Durable mailboxes are superseded by akka-persistence
, which offers several
tools to support reliable messaging.
Read more about akka-persistence
in the documentation for Scala and
documentation for Java.
Deprecated STM Support for Agents
Agents participating in enclosing STM transaction is a deprecated feature.
Transactor Module is Deprecated
The integration between actors and STM in the module akka-transactor
is deprecated and will be
removed in a future version.
Typed Channels has been removed
Typed channels were an experimental feature which we decided to remove: its implementation relied on an experimental feature of Scala for which there is no correspondence in Java and other languages and its usage was not intuitive.
Removed Deprecated Features
The following, previously deprecated, features have been removed:
- event-handlers renamed to loggers
- API changes to FSM and TestFSMRef
- DefaultScheduler superseded by LightArrayRevolverScheduler
- all previously deprecated construction and deconstruction methods for Props
publishCurrentClusterState is Deprecated
Use sendCurrentClusterState
instead. Note that you can also retrieve the current cluster state
with the new Cluster(system).state
.
CurrentClusterState is not a ClusterDomainEvent
CurrentClusterState
does not implement the ClusterDomainEvent
marker interface any more.
Note the new initialStateMode
parameter of Cluster.subscribe
, which makes it possible
to handle the initial state as events instead of CurrentClusterState
. See
documentation for Scala and
documentation for Java.
BalancingDispatcher is Deprecated
Use BalancingPool
instead of BalancingDispatcher
. See documentation for Scala and
documentation for Java.
During a migration period you can still use BalancingDispatcher by specifying the full class name in the dispatcher configuration:
type = "akka.dispatch.BalancingDispatcherConfigurator"
akka-sbt-plugin is Removed
akka-sbt-plugin
for packaging of application binaries has been removed. Version 2.2.3 can still be used
independent of Akka version of the application. Version 2.2.3 can be used with both sbt 0.12 and 0.13.
sbt-native-packager is the recommended tool for creating distributions of Akka applications when using sbt.
Parens Added to sender
Parens were added to the sender()
method of the Actor Scala API to highlight that the sender()
reference is not referentially transparent and must not be exposed to other threads, for example by closing over it when using future callbacks.
It is recommended to use this new convention:
sender() ! "reply"
However, it is not mandatory to use parens and you do not have to change anything.
ReliableProxy Constructor Changed
The constructor of ReliableProxy
in akka-contrib
has been changed to take an ActorPath
instead of
an ActorRef
. Also it takes new parameters to support reconnection. Use the new props factory methods, ReliableProxy.props
.
OSGi Changes
akka-osgi
no longer contains the akka-actor
classes, instead
akka-actor
is a bundle now. akka-osgi
only contains a few OSGi helpers,
most notably the BundleDelegatingClassLoader
which resolves e.g.
reference.conf
files (hence these are not copied into the akka-osgi
bundle any longer either).
akka-osgi-aries
has been removed. Similar can be implemented outside of Akka if needed.
TestKit: reworked time dilation
TestDuration
has been changed into an implicit value class plus a Java API in JavaTestKit. Please change:
import akka.testkit.duration2TestDuration
into:
import akka.testkit.TestDuration
Contents