* Needed in Lagom to be able to have separate object mappers with different
config for exernal and internal usage.
* Can also be good to be able to have different config for json and cbor
serializers, or different for remote messages and persisted events.
* Pass in binding name when creating the serializer if it has a matching
constructor
* Serialization identifiers loaded from config via the binding name instead
of class name, for JacksonSerializer.
* Decrease waiting-for-state-timeout to 2s
The timeout can be likely to happen for small clusters when the read
majority needs to go to every node
Refs #26710
* Restructure migration guide for 2.6
Separate out changes that require no user action unless the old
behavior is preferred
* Update akka-docs/src/main/paradox/project/migration-guide-2.5.x-2.6.x.md
Co-Authored-By: Arnout Engelen <github@bzzt.net>
* previous `schedule` method is trying to maintain a fixed average frequency
over time, but that can result in undesired bursts of scheduled tasks after a long
GC or if the JVM process has been suspended, same with all other periodic
scheduled message sending via various Timer APIs
* most of the time "fixed delay" is more desirable
* we can't just change because it's too big behavioral change and some might
depend on previous behavior
* deprecate the old `schedule` and introduce new `scheduleWithFixedDelay`
and `scheduleAtFixedRate`, when fixing the deprecation warning users should
make a concious decision of which behavior to use (scheduleWithFixedDelay in
most cases)
* Streams
* SchedulerSpec
* test both fixed delay and fixed rate
* TimerSpec
* FSM and PersistentFSM
* mima
* runnable as second parameter list, also in typed.Scheduler
* IllegalStateException vs SchedulerException
* deprecated annotations
* api and reference docs, all places
* migration guide
* Netty is not optional for multi jvm test kit
Netty is used internally and is always loaded even if artery is enabled.
That could be improved but for now netty is needed.
This fixes the Akka HTTP against master build
This doesn't change the fact that the FJP in JDK 11 is less fair but makes the test pass:
* Disable serialize messages to make each message processing less CPU-intense
* Tune the throughput down to the default (was a very high value, 100), so less messages are worked through in one batch of internal dispatch.
* most convenient for users to include it akka-serialization-jackson
and load it when akka-actor-typed is in classpath
* provided dependency to akka-actor-typed
* Harden RemotingSpec: Retry if address is taken
The test requires that the ActorSystem not be started when doing the
selection to using port = 0 is not an option.
Fixes#25657
