NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
Maven Central, #951 by repository stars
Reactor Test support
Latest stable
3.9.0
Last release 2 days ago
20 Nov 2026
Ships fairly regularly
a new release about every 5 weeks
Nearly every release is documented
notes for 59 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
7 years old
185 releases · first in 2019
Drag across the ruler to compare, or tap a release to open it.
fix #2468 Review and polish deprecation suppressions
Reactor-Core 3.4.1 is part of 2020.0.2 Release Train (Europium SR2).
This service release contains bugfixes and new features.
This note focuses on changes specific to this version, but all changes from 3.2.22.RELEASE and 3.3.12.RELEASE are also included.
Timed wrapper object, allowing richer output of Duration compared to elapsed() and timestamp() (which use Tuple2 and Long)LoggerUtils allows the early installation of a Logger factory that will forward to a classic factory AND let the user activate a logging indirection later on during testsTestLogger to capture internal logs from operators that were classloaded earlier than the unit testTestPublisher
createColdNonCompliant variantContext in StepVerifier, say from which operator the asserted Context was taken (#2518)RetryWhenMainSubscriber#onError's calls serial (#2499, Fixes #2488)@dowgiallom, @seants
Your coding agent can read these notes before it upgrades. Set up the MCP server →
Schedulers deprecated back in 3.3 have been removed (WorkQueueProcessor and TopicProcessor, #2208)
Reactor-Core 3.4.0 is part of 2020.0.0 Release Train (codename Europium).
This is the first GA release of Europium :tada:
This note focuses on 3.4.0 proper, curating changes from across all milestones.
Note that the 3.4.0 effort was started right after 3.3.5.RELEASE, so this version also contains changes from 3.3.6 to 3.3.11 (see the end of the notes for links).
firstWithValue new operator has an undocumented way of checking the cause of failure for individual sources, but it uses a suppressed exception rather than a cause. Please refrain from using that, as we'll switch to initCause in 3.4.1 (and properly document the behavior, setting it in stone) => #2459WorkQueueProcessor and TopicProcessor, #2208)Schedulers.elastic() and variants are deprecated, to be removed in 3.5 (#1893)FluxProcessor (as well as all concrete implementations of it) are now deprecated and will be removed in 3.5 (#2188, #2431)
MonoProcessor is made abstract and deprecated, to be removed in 3.5 (#2188, #1053 / f04bede3, #2431)
Flux.subscribe(Consumer<Subscription>) variant is deprecated, to be removed in 3.5 (#1825)first is deprecated to be renamed to firstWithSignal (see firstWithValue in the new features below). The old deprecated method will be removed in 3.5 (#2173)Context changes (#2282):
Context at points where there is no use attempting to use write methods are deprecated and will be removed in 3.5 (#2293, 7d6c8625)
ContextView-based variant (see new features below)deferWithContext => deferContextual with a ContextViewSignal.getContext() => Signal.getContextView()Context are renamed to be clearer. Old names are deprecated and will be removed in 3.5 (see new feature below, #2148, eed93d64)Duration and convert it to milliseconds now use nanosecond precision (#1734)Operators.onErrorDropped no longer throws, but only logs the dropped exception (#1431)onError signal reaches the last (lambda based) subscriber in a chain which doesn't have an error handler, the exception is no longer thrown but only logged (#2176)Scheduler now initiate their backing infrastructure and apply Schedulers decorator in start() rather than the constructor (#1778)
start() is also now implicitly called when instantiating a Scheduler via the Schedulers.Factorythis leaking from constructorstart()EmitterProcessor with a full backpressure queue no longer blocks until data is requested, but fails fast (#2049)name operator (#1928, #2302)
flow tag has been removed entirelyname({String}) operator will replace reactor. prefix in meter names with the user-provided {String}.tags(String...) operator, but not name(String) still creates a sparse set of tags in the application which could be rejected by some metrics backendreactor_requested_requested_amount => {name}_requestedreactor_subscribed_subscribers => {name}_subscribed{name} into account from the above changestatus tag, introducing completedEmpty value (#1803)SinksSinks are a new API constructed to replace the FluxProcessor and MonoProcessor APIs, which have been deprecated (#2431, #1053).
We distinguish Sinks.Many<T> (a Flux-like sink), Sinks.One<T> (a Mono-like sink) and Sinks.Empty (a Mono<Void>-like sink).
These sinks are constructed through specs under the Sinks class. They expose a two-faced API:
tryEmit are atomic, never throw (#2319, #2329, #2336, #2426) and immediately return an EmitResult enum (originally named Emission) indicating various variations of success or error (#2319, #2338)emit attempt to provide an easier facade over the above, take an EmitFailureHandler allowing to fine tune some aspects of the default behavior (#2377)By default, the instances are "serialized", detecting parallel usage and failing fast in tryEmit (#2342, #2410, #2365, 19fc1ba1, #2412).
The vanilla reactor sink implementations of emit API terminate with the equivalent of an emitError when parallel usage is detected (#2365) but by using the EmitFailureHandler to drive the sink to retry, possibly with a small amount of sleeping, it should be possible to optimistically get rid of the contention.
It is also possible to get a low-overhead, low-protection instance of vanilla sinks by using the Sinks.unsafe() spec (#2418).
This doesn't detect parallel usage, so it must be used in a responsible way (in a context where the Reactive Streams contract is already externally enforced).
Sinks also expose a asFlux() view (or asMono(), as relevant) for them to be passed to downstream and subscribed, and also expose a bit of state through currentSubscriberCount() getter (#2372) and by being Scannable (at a minimum for TERMINATED and CANCELLED attributes, #2394).
Compared to existing processors, most flavors have a sink equivalent except DirectProcessor, with two close cousins under Sinks.many().multicast() (#2392, #2451):
directAllOrNothing will ignore tryEmitNext attempts when at least one subscriber doesn't have enough demand,directBestEffort will instead push to the subset of subscribers with demand and drop from the perspective of slow subscribers.This differs from DirectProcessor in the sense that both these sinks are kept open in such a situation, so slow subscribers that increase their request will start seeing values pushed after that (instead of being immediately terminated in DirectProcessor).
There is also a new flavor: an onBackpressureError() variant of the unicast Sink (#2347)
See links to Dysprosium releases (3.3.6 to 3.3.11) at the end.
concatMap now has a variant with 0 prefetch (#2202)Scannable operators answer the new Attr.RUN_STYLE attribute with a RunStyle enum (#2058, #2123)
Publisher and Subscriber levelRunStyle.SYNC guarantees the operator doesn't change threadsRunStyle.ASYNC indicates the operator MAY change threadsRunStyle.UNKNOWN (the default) indicates there's no available information (eg. due to the operator wrapping an arbitrary Publisher)MetricsRegistry other than the global one can now be configured (#2253, 64fd556)
Metrics.MicrometerConfiguration#useRegistry for that purposeContext now extends a simplified ContextView API which only expose the read methods (#2279, #2293)
Context when the only meaningful operations are read operationsSignal#getContextView(), deferContextual operator (variant to deferWithContext) and new transformDeferredContextual operatortransformDeferredContextual operator that takes a BiFunction, allowing to access the ContextView in the transformation (#2280 then renamed in #2293)
Mono.subscribe(valueConsumer, errorConsumer, ...) error Consumer now gets notified of exceptions thrown by the value Consumer (#1995)Context operators with clearer names (#2148, eed93d64)
Mono.subscriberContext() is not aliased, we now advise to use Mono.deferContextual(Mono::just) to get the exact same behaviorsubscriberContext(Context) is replaced by contextWrite(ContextView) (notice the use of ContextView)contextWrite(Function)Schedulers.setFactoryWithSnapshot(Factory) to get a Snapshot object while replacing the FactoryresetFrom(Snapshot)VirtualTimeScheduler to reset previously customized factoriesParallelFlux.subscribe(array) method is now public to allow delegation in wrappers (#2328)Retry object driving #retryWhen operator can now store user-provided state in the form of a ContextView, which is exposed through RetrySignal (#2312, bd8db8ab)GroupedFlux#key() method is now marked as @NonNull (#2397)firstWithValue factory operator (#2173, 41c937ff)
first, it let multiple sources competeNoSuchElementException is thrownlog() will log context access at FINE(ST) level and now uses the more correct currentContext prefix (#2220)Mono#materialize() now takes into account that the source IS a Mono, so it doesn't cancel() it when materializing its onNext (#2424, 765bfe8f)
@AayushyaVajpayee, @camsteffen, @cnabro, @hamidshahid, @Inego, @jonenst, @josemalonsom, @OlegDokuka, @robotmrv, @seants, @smaldini, @steppedreckoner, @yschimke
All changes from these releases have been forward-merged into 3.4.0:
This is the last release of 3.3.x, as the Dysprosium line is now EoL
This is the last release of 3.3.x, as the Dysprosium line is now EoL
@NonNull annotation, not JetBrain's NotNull by @simonbasle in #2808Full Changelog: https://github.com/reactor/reactor-core/compare/v3.3.21.RELEASE...v3.3.22.RELEASE
Reactor-Core 3.3.21 is part of `Dysprosium-SR24` Release Train.
Reactor-Core 3.3.21 is part of Dysprosium-SR24 Release Train.
This service release contains one bugfixe and dependency bumps (including Gradle 7.2).
Reactor-Core 3.3.20.RELEASE is part of `Dysprosium-SR23` Release Train.
Reactor-Core 3.3.20.RELEASE is part of Dysprosium-SR23 Release Train.
This service release contains bugfixes and polishing.
FluxOnBackpressureBufferStrategy reject sizes <= 0 (#2757)@hisener
Reactor-Core 3.3.19.RELEASE is part of `DYSPROSIUM-SR22` Release Train.
Reactor-Core 3.3.19.RELEASE is part of DYSPROSIUM-SR22 Release Train.
This service release contains a bugfix, as well as some build/ci polishing.
spotless plugin introduced to check java headers (#2722, #2732)Reactor-Core 3.3.18.RELEASE is part of `Dysprosium-SR21` Release Train.
Reactor-Core 3.3.18.RELEASE is part of Dysprosium-SR21 Release Train.
This minor service release contains a few test and build enhancements.
Reactor-Core 3.3.17.RELEASE is part of `Dysprosium-SR20` Release Train.
Reactor-Core 3.3.17.RELEASE is part of Dysprosium-SR20 Release Train.
This service release contains a few bugfixes and enhancements.
main, update CI (#2679)Reactor-Core 3.3.16.RELEASE is part of `Dysprosium-SR19` Release Train.
Reactor-Core 3.3.16.RELEASE is part of Dysprosium-SR19 Release Train.
This service release contains bugfixes and a couple implementation improvements.
@bruto1
Reactor-Core 3.3.15.RELEASE is part of `Dysprosium-SR18` Release Train.
Reactor-Core 3.3.15.RELEASE is part of Dysprosium-SR18 Release Train.
This service release mainly contains bugfixes.
MonoInnerProducerBase
MonoInnerProducerBase (#2632)MonoProducerBase but remove fusion support from MonoSingle (#2645)Mono.fromSupplier(() -> null).single() previously failed to produce an onError(NoSuchElementException) (#2635)
Mono that was implementing Callable (which includes fromSupplier, fromCallable, fromRunnable to some extent...) would exhibit the wrong behaviorFlux#single, which is going to be fixed by #2657 in the next release@martin-tarjanyi
Reactor-Core 3.3.14.RELEASE is part of `Dysprosium-SR17` release train.
Reactor-Core 3.3.14.RELEASE is part of Dysprosium-SR17 release train.
This is a maintenance release.
skipUntil conditional subscriber logic inconsistent (#2578)concatMap (#2576)@koldat, @mickeelm, @seants, @sjh836
Reactor-Core 3.3.13.RELEASE is part of `Dysprosium-SR16` Release Train.
Reactor-Core 3.3.13.RELEASE is part of Dysprosium-SR16 Release Train.
This service release contains a couple bugfixes and documentation polishing.
@OlegDokuka
Reactor-Core 3.3.12.RELEASE is part of `Dysprosium-SR15` Release Train.
Reactor-Core 3.3.12.RELEASE is part of Dysprosium-SR15 Release Train.
This service release contains a couple bugfixes and documentation polishing.
This note focuses on changes specific to this version, but all changes from 3.2.22.RELEASE are also included.
Mono contract was not sufficiently and explicitly stated in the docs, this has been fixed (#2481)@BilliAlpha
fix #2416 Don't blanket hide deprecated members in javadoc
Reactor-Core 3.3.11.RELEASE is part of Dysprosium-SR13 Release Train.
This is a recommended update for all Reactor 3 users.
This release also contains all changes available in 3.2.21.RELEASE
@yosfik
Reactor-Core 3.3.10.RELEASE is part of `Dysprosium-SR12` Release Train.
Reactor-Core 3.3.10.RELEASE is part of Dysprosium-SR12 Release Train.
This maintenance release contains one small enhancement to DirectProcessor/EmitterProcessor.
This version also contains all changes from v3.2.20.RELEASE.
Nothing published for this version
Reactor-Core 3.3.8.RELEASE is part of `Dysprosium-SR10` Release Train.
Reactor-Core 3.3.8.RELEASE is part of Dysprosium-SR10 Release Train.
This maintenance release is a recommended update for all Reactor 3 users.
This version also contains all changes from v3.2.19.RELEASE.
null publisher in MonoDelayUntil (#2245)@dennyac, @nicolasgavalda
reactor-core 3.3.7.RELEASE is part of `Dysprosium-SR9` Release Train.
reactor-core 3.3.7.RELEASE is part of Dysprosium-SR9 Release Train.
Flux.collect() (#2181)@rstoyanchev , @melix , @gurkerl83
Reactor-Core 3.3.6.RELEASE is part of `Dysprosium-SR8` Release Train.
Reactor-Core 3.3.6.RELEASE is part of Dysprosium-SR8 Release Train.
DeferredSubscription makes requests serially if there is racing between set / requestN (according to RS spec Rule 2.7)(#2164)FluxSwitchOnFirst serializes requests (according to RS spec Rule 2.7) (#2161)UnicastProcessor disposed to avoid racing between onNext and dispose (#2180)ZoneId.of(String) to avoid blocking disk load during first access of Schedulers.boundedElastic() (#2143)ASYNC fusion in Mono.metrics() to avoid NullPointerException or hanging in certain cases (#2160 )ConditionalSubscriber classcast in FluxSwitchOnFirst (#2166)Subscription set by reportThrowInSubscribe to an instance compatible with QueueSubscription, preventing fused cases where a ClassCastException would be thrown. (#2152 )publishOn scan test, hiccup before asserting BUFFEREDMono.first() and Mono.or() behavior (#2159)Retry.fromFunction() instead of from() (#2172)RetrySpec doc: throwablePredicate is actually filter (#2162)onCancel / onDipose limited to 1 use (#2015)generate code example (#2110)@alex-dukhno , @chakas , @berry120 , @OlegDokuka , @PyvesB , @josemalonsom
reactor-core 3.3.5.RELEASE is part of `Dysprosium-SR7` Release Train.
reactor-core 3.3.5.RELEASE is part of Dysprosium-SR7 Release Train.
It also contains all fixes from 3.2.17.RELEASE.
subscribe method can lead to two onSubscribe signal if said exception is thrown after having sent the Subscription (#2107, #2111, #2072)StepVerifier cannot, which can "hide" the true errorSubscription can be identified with Operators.canAppearAfterOnSubscribe(Subscription), which is used by StepVerifier to ignore that particular extra subscriptionsubscribe/subscribeOrReturn mutualized implementation, prevent StepVerifier getting confused by a double onSubscribe in that very specific case ((#2072, #2107, #2111)onNext vs cancel races that would bypass resource discarding (#2134, #2126, #2102, #2124, #2125, #2122, #2077, #2102)
MonoDelayElement#onNext downstreamerrorOnSubscribe steps (#2112)errorOnSubscribe scenarios into separate tests (#2099)@TestFactory in BaseOperatorTest (#2085)@OlegDokuka, @pimlock
retryWhen(Function), all retryBackoff and retry(Predicate)/retry(long, Predicate) are now deprecated, to be replaced by the more generic retryWhen(Ret…
reactor-core 3.3.4.RELEASE is part of Dysprosium-SR6 Release Train.
retryWhen(Function), all retryBackoff and retry(Predicate)/retry(long, Predicate) are now deprecated, to be replaced by the more generic retryWhen(Retry) (see RetrySpec and RetryBackoffSpec for configurable builders that cover these methods and more)retryWhen(Function) and retryBackoff as well as some retry variants in 3.4.0Retry abstract class, with two builders (or Spec) to fluently configure the retriesonNext comes in between errorsrepeatWhen and retryWhen, when a Context is emitted, we now merge it with the current instead of replacing it (#2056)VirtualTimeScheduler#getOrSet(boolean) fixes virtual time in the StepVerifier (#2060, #2017)@OlegDokuka
reactor-core 3.3.3.RELEASE is part of `Californium-SR16` Release Train.
reactor-core 3.3.3.RELEASE is part of Californium-SR16 Release Train.
defer opt-in in VirtualTimeScheduler.create (#1251, #2012)
VirtualTimeScheduler will now eagerly advance time even if there is no pending task. Use the overload with defer=true to get the old default behavior.StepVerifier-created ones will continue lazily advancing time when no task is scheduled, in order to make testing of things like delays inside a flatMap possible.fromStream and fromIterable with custom non-Collection Iterables, see #2014 and #2021
iterator() method is now invoked twice per wrapping of a given Iterable (once through default spliterator() to check the iterable is finite aka. hasCharacteristic(SIZED), once for the iteration itself)Iterator is discarded when cancelled mid-iterationIterable is a Collection or its Spliterator has the SIZED characteristic.Iterable#iterator for custom iterablesCallable" in Flux#collectList (#2023, #2024)@btkelly, @ilyavy, @robotmrv, @yarosla
reactor-core 3.3.2.RELEASE is part of `Dysprosium-SR3` Release Train.
reactor-core 3.3.2.RELEASE is part of Dysprosium-SR3 Release Train.
Flux.replay/cache hanging or delivering wrong data (tracked in #2030)currentContext() with ArchUnit (#1988)@echooymxq, @hojongs, @martin-tarjanyi, @mdonkers, @OlegDokuka, @sgsfak
Reactor-Core 3.3.1.RELEASE is part of `Dysprosium-SR2` Release Train.
Reactor-Core 3.3.1.RELEASE is part of Dysprosium-SR2 Release Train.
It also includes changes from the 3.2.13.RELEASE version.
One important change to pick up for testing Reactor is in StepVerifier: the combination of expectNoEvent followed by thenCancel wasn't working terribly well, and a new better terminal expectation has been introduced: expectTimeout (#1913)
Flux.replay/cache hanging or delivering wrong data (tracked in #2030)StepVerifier: expectTimeout(Duration) as an alternative to expectNoEvent(Duration).thenCancel()transform or transformDeferred apply a transformation to a non-Reactor Publisher, warn when/if the original Context is lost (#1888)collectSortedListMap.Entry in Context1 to ensure various libraries don't represent it inconsistenly when printing out collections of ContextContextN (by extending LinkedHashMap instead of HashMap)Mono#cache(Function, Function, Supplier) (backported)VirtualTimeScheduler#getScheduledTaskCount() (backported)Flux#publish() propagate 1st subscriber ContextExceptions.isTraceback, and a mean to unwrap a composite without the tacebacks (Exceptions.unwrapMultipleExcludingTracebacks)parallel() not replenishing properlybufferWhile was incorrectly showing dropped separatorsFlux#join javadoc and marble have been clarifiedmaven-publish and artifactory (#1943)Exceptions.multiple(Iterable) now produces a CompositeException, more test coverage of Exceptions util (#1954)@andreisilviudragnea, @akarnokd, @berry120, @dittos, @EleanorRoseLegg, @gindex, @jabrena, @simondaudin, @slisaasquatch, @thiyagu-7, @ttddyy, @slisaasquatch, @wonwoo
Mono#do[On|After]SuccessOrError has been deprecated and marked for removal in 3.4
Reactor-Core 3.3.0.RELEASE is part of Dysprosium-RELEASE Release Train.
This release note covers the first GA release of the 3.3.x cycle, and as such describes the changes from the perspective of a 3.2.x to 3.3.0 comparison.
See also release notes from 3.2.12, 3.2.11, 3.2.10 and 3.2.9 for changes that were inherited from the 3.2.x maintenance releases.
This release is a big opportunity to polish the APIs and some behaviors with less restriction than maintenance releases.
🌟 It also introduces a few new features, the highlight of which is the BoundedElasticScheduler:
this one has evolved from the ElasticScheduler and is intended as a future replacement for it.
A cap can be put on the number of threads such a scheduler can spawn. After that cap is reached, further task submissions are deferred (being enqueued) up to a second global limit on enqueued tasks (optionally unbounded).
Flux.replay/cache hanging or delivering wrong data (tracked in #2030)onNext: they could trigger onError after onNext, which is not permitted in Mono.
Mono#usingWhen now waits for the async complete handler termination before propagating onNext (#1832)Mono#using similarly avoids triggering onNext+onError if cleanup fails (#1853)Mono#doOnTerminate now acts similarly to doOnSuccess, executing the side effect before the onNext is propagated downstream (#1752)Mono#do[On|After]SuccessOrError has been deprecated and marked for removal in 3.4 (#1854)usingWhen API has been polished, and some variants deprecated for removal in 3.4 (#1687)
BiFunction for error handler (exposing the failure that causes "rollback")Mono for alignment with Flux. This is the simplest case: same handler for complete, error and cancel terminations.WorkQueueProcessor and TopicProcessor have been marked as deprecated for 3.3.x and will be removed in 3.4UnsafeEmitterProcessor and publishOnreactor-kotlin-extensionsSchedulers.Factory#decorateExecutorService has been removedonBackpressureBuffer(maxSize+error) variants now delay errorscompose() has been deprecated and renamed transformDeferred()SignalonNext events to be emitted only when the subsequent signal is ready upstreamonComplete signal, the corresponding Signal must be explicitly requested from upstream (one extra request(1))Exceptions#unwrap) in most sources that immediately produce an error, acting like a Callable (b6f4d291)reactor-tools sources into the core repository (#1731)boundedElastic Scheduler that caps number of threads and scales to 0 (#1804)
scan(Attr.NAME) or toString) of global singleton Schedulers (#1885, #1886)
parallel(parallel) this is now Schedulers.parallel()Schedulers.DEFAULT_POOL_SIZE (which can be overridden by system property)subscribe from stack by optimizing the "tail calls" (#1761, #1826)Context lookups in error sources (4f2bc7a7)DelayedWorkQueue#take (#1726)Automatic-Module-Name (reactor.core and reactor.test, #1641, #1692)CorePublisher interface has been introduced to distinguish Publisher from reactor (#1527)toString representation of classes that are known to have an obscure representation, as well as extracting such objects that would be nested in containers like a Collection, Tuple2, Signal, ...Throwable as a suppressed exception@aftersss, @akarnokd, @apuravchauhan, @berry120, @Buzzardo, @dittos, @don-vip, @flambard, @gindex, @hisener, @izeye, @jnizet, @mumukiller, @OlegDokuka, @pcdavid, @pmackowski, @ramonPires, @szpak, @ttddyy, @UgiR, @zhouzhou19950825
Reactor-Core 3.2.22.RELEASE is part of `Californium-SR23` Release Train.
Reactor-Core 3.2.22.RELEASE is part of Californium-SR23 Release Train.
This is the last release of Californium 👋
This service release contains a bug fix, as well as some polishing changes.
@koldat
Reactor-Core 3.2.21.RELEASE is part of `Californium-SR22` Release Train.
Reactor-Core 3.2.21.RELEASE is part of Californium-SR22 Release Train.
This is a minor maintenance release polishing the build and documentation.
@yosfik
[build] Various polish of Gradle build (deprecation warnings #2274, wrapper validation #2348)
Reactor-Core 3.2.20.RELEASE is part of Californium-SR21 Release Train.
This is a maintenance release with a couple bugfixes, as well as build polish.
@seants, @sullis
Reactor-Core 3.2.19.RELEASE is part of `Californium-SR20` Release Train.
Reactor-Core 3.2.19.RELEASE is part of Californium-SR20 Release Train.
This is a maintenance release with a couple bugfixes, as well as build polish.
@arturdryomov, @gurkerl83, @rstoyanchev
Reactor-Core 3.2.18.RELEASE is part of `Californium-SR19` Release Train.
Reactor-Core 3.2.18.RELEASE is part of Californium-SR19 Release Train.
DeferredSubscription makes requests serially if there is racing between set / requestN (according to RS spec Rule 2.7)(#2164)UnicastProcessor disposed to avoid racing between onNext and dispose (#2180)@OlegDokuka
reactor-core 3.2.17.RELEASE is part of `Californium-SR18` Release Train.
reactor-core 3.2.17.RELEASE is part of Californium-SR18 Release Train.
onNext vs cancel races that would bypass resource discarding (#2134, #2126, #2102, #2124, #2125, #2122, #2077, #2102)
MonoDelayElement#onNext downstream@OlegDokuka, @pimlock
reactor-core 3.2.16.RELEASE is part of `Californium-SR17` Release Train.
reactor-core 3.2.16.RELEASE is part of Californium-SR17 Release Train.
reactor-core 3.2.15.RELEASE is part of `Californium-SR16` Release Train.
reactor-core 3.2.15.RELEASE is part of Californium-SR16 Release Train.
fromStream and fromIterable with custom non-Collection Iterables, see #2014 and #2021
iterator() method is now invoked twice per wrapping of a given Iterable (once through default spliterator() to check the iterable is finite aka. hasCharacteristic(SIZED), once for the iteration itself)Iterator is discarded when cancelled mid-iterationIterable is a Collection or its Spliterator has the SIZED characteristic.Iterable#iterator for custom iterablesCallable" in Flux#collectList (#2023, #2024)@btkelly, @robotmrv
Nothing published for this version
Reactor-Core 3.2.13.RELEASE is part of `Californium-SR14` Release Train.
Reactor-Core 3.2.13.RELEASE is part of Californium-SR14 Release Train.
Mono#cache(Function, Function, Supplier) (backported)VirtualTimeScheduler#getScheduledTaskCount() (backported)Flux#publish() propagate 1st subscriber ContextExceptions.isTraceback, and a mean to unwrap a composite without the tacebacks (Exceptions.unwrapMultipleExcludingTracebacks)parallel() not replenishing properlymaven-publish and artifactory (#1943)Exceptions.multiple(Iterable) now produces a CompositeException, more test coverage of Exceptions util (#1954)@andreisilviudragnea, @akarnokd, @ttddyy
Reactor-Core 3.2.12.RELEASE is part of `Californium-SR11` Release Train.
Reactor-Core 3.2.12.RELEASE is part of Californium-SR11 Release Train.
This is a recommended update for all Reactor 3 users.
Flux#flatMap with scalar source now supports onErrorContinue (#1684)Flux#doAfterTerminate now executed when onError in fused mode (#1836, #1835)LinkageError in Traces. This helps with some static analyzers (#1830)cancelSupport when lifting ConnectableFlux (#1860)CONTRIBUTING.md and issue_template in favor of centralized ones (740c0483)@gindex
Reactor-Core 3.2.11.RELEASE is part of `Californium-SR10` Release Train.
Reactor-Core 3.2.11.RELEASE is part of Californium-SR10 Release Train.
This is a recommended update for all Reactor 3 users, with mainly bugfixes.
null as a name@gindex
Reactor-Core 3.2.10.RELEASE is part of `Californium-SR9` Release Train.
Reactor-Core 3.2.10.RELEASE is part of Californium-SR9 Release Train.
This is a recommended update for all Reactor 3 users.
Scheduler for retryBackoff (#1707)Mono#onErrorStop() convenience API to align with Flux (#1728)TestPublisher assertion method assertMaxRequested (#1732)doFirst operator provides a clearer dual to doFinally, as a better alternative to common usage of doOnSubscribe (#1716)FluxPublish#remove() (2ce7b75f)Flux.publish().autoConnect(0) now correctly drops early items (#1712)Mono#doOnSuccessOrError (#1725)@don-vip
Reactor-Core 3.2.9.RELEASE is part of `Californium-SR8` Release Train.
Reactor-Core 3.2.9.RELEASE is part of Californium-SR8 Release Train.
This is a recommended update for all Reactor 3 users.
maxSize (#1666)@aftersss, @OlegDokuka
Reactor-Core 3.2.8.RELEASE is part of `Californium-SR6` Release Train.
Reactor-Core 3.2.8.RELEASE is part of Californium-SR6 Release Train.
It is an immediate fixup release, fixing a regression introduced in 3.2.7.RELEASE (3d2d1c40).
The rest of this release note reflects the content of 3.2.7.RELEASE, that is to say the delta with 3.2.6.RELEASE
onCancel handler on eager cancellation of Flux.create and Mono.create (#1622, 3d2d1c40)FluxBufferTimeout on flush (#1534)onErrorContinue needs idempotent predicate, mention throwing (c15ff917)This release contains a regression, v3.2.8.RELEASE was released immediately after. Prefer using the later.
This release contains a regression, v3.2.8.RELEASE was released immediately after. Prefer using the later.
Reactor-Core 3.2.6.RELEASE is part of `Californium-SR5}` Release Train.
Reactor-Core 3.2.6.RELEASE is part of Californium-SR5} Release Train.
This is a recommended update for all Reactor 3 users.
ConsoleLogger doesn't log TRACE/DEBUG levels by default now (#1484)
Loggers.useVerboseConsoleLoggers()reactor.logging.fallback system property to JDKFlux.collect and friends (collectList, collectMap, collectMultiMap) now support doOnDiscard (#1510)Scheduler#schedulePeriodically now accepts 0 delay and 0 duration periodic tasks (#1498)onNext(QueueSubscription) (#1515)
UnsupportedOperationException to detect cases of attempting to "deep-logging" an internal object as a CollectionQueueSubscription) to a String before passing to logging implementation.switchOnFirst has been improved
Publisher (#1517)Flux will get cancelled at the termination of said unrelated publisherBiFunction (#1502)Mono#fromRunnable can now be cancelled immediately and withholds invoking the Runnable (#1503)windowWhile/windowUntil cancellation when the resulting Flux is cancelled before the last in flight window (#1452)mergeSequential's first inner sequence would not be cancelled with the rest when output Flux was cancelled (#1500)usingWhen terminal callbacks are now mutually exclusive with the cancel one (#1487)index,limitRequest,onBackpressureBuffer variant, sort, sequenceEquals, doFinally, doOnEach, switchOnFirstswitchOnFirst has been clarified, regarding the inclusion of first signal in the Flux passed to the BiFunction (#1514)reactive-stream Gradle coordinates in reference documentation (#1505)@glandais-eptica, @OlegDokuka
This is the 6th release of Reactor 3.2, part of `Californium-SR4` Release Train.
This is the 6th release of Reactor 3.2, part of Californium-SR4 Release Train.
This is a recommended update for all Reactor 3 users, containing a few improvements, a new operator (switchOnFirst) and documentation improvements (the biggest being a switch from PNG marbles to SVG marbles).
doc-files) means we don't rely on being online to fetch and display them.@davinkevin, @fletchgqc, @mlevvy, @npathai, @OlegDokuka, @pavelhoral, @sdeleuze, @thekalinga
It has been built using JDK11 and the javadoc jars are non-functional, especially when served from the reactor website.
v3.2.5.RELEASE-It has been built using JDK11 and the javadoc jars are non-functional, especially when served from the reactor website.
See release notes at https://github.com/reactor/reactor-core/releases/tag/v3.2.5.RELEASE
This replaces the Factory method that is now deprecated (but applied as the last decorator)
This is the 4th release of Reactor 3.2, part of Californium-SR3 Release Train.
This is a recommended update for all Reactor 3.2 users.
Factory method that is now deprecated (but applied as the last decorator)String, allowing for easy identification and removal of specific decorators and good behavior when multiple libraries need to install a decoratorMono.fromFuture/Mono.fromCompletionStage now correctly detects when the future is completed AFTER the Mono has been cancelled (#1414)
CancellationExceptionFuture itself in a doOnCancelScheduler backed by an ExecutorService can now be instrumented with Micrometer (#1201)Schedulers.enableMetrics() method enables thisScheduler implementations can apply the same decorator (and any registered decorator) via the new Schedulers.decorateExecutorService methodexpectAccessibleContext work with initialContext from StepVerifier when just asserting a scalar source without operators (#1417)onBackpressureLatest (#1430)Scheduler#createWorker interface has been clarified not to always include ordering (#1421)distinctUntilChanged*DoesntRetainObjects (#1428)@nurkiewicz @lerencao @pranavek @quaff @OlegDokuka
This is the 3rd release of Reactor 3.2, part of `Californium-SR2` Release Train.
This is the 3rd release of Reactor 3.2, part of Californium-SR2 Release Train.
This maintenance release contains a few bugfixes, as well as minor improvements.
Violation.DEFER_CANCELLATION (#1406)reduce, prevent value retention in scan (#1394)MonoReduce rely on super.value, null it out in complete (#1410)metrics() without Micrometer triggers NoClassDefFoundError, now fixed (#1383)share() and .publish().refCount() (including one with grace period) missing repeat/retry re-subscribes (#1385, #1386)reduce() lambda sometimes being passed a null value, potentially causing NPE (#1393)finalize in test by introducing PhantomReference-based tracker in new MemoryUtils (#1407)@philleonard, @zhouzhou19950825
This is the second release of Reactor 3.2, part of `Californium-SR1` Release Train.
This is the second release of Reactor 3.2, part of Californium-SR1 Release Train.
This release contains a couple of small bugfixes, recommended for all 3.2 users, as well as several minor improvements to the javadoc/reference documentation.
/docs/, not /src (#1360, f0fd5591)EmitterProcessor SYNC fusion completing too early, not producing an element when fused (#1364)CombineLatest ClassCastException with single source(#1356)@nurkiewicz, @pgilad, @romangr
(deprecation) Scannable#operatorName() is deprecated in favor of stepName() (#1115, #1140)
This is the first release of Reactor 3.2, part of the Californium-RELEASE Release Train.
This major release contains a few significant new features, in addition to a heap of improvements and bug fixes.
3.1.x, the Scannable#operatorName() method has been renamed to stepName() (#1156)Disposables.composite() now has list semantics rather than undocumented set semantics, and thus should be more performant (#1237)Schedulers.DEFAULT_POOL_SIZE is now used internally (notably by Schedulers.[new]parallel()), but unlike before it doesn't enforce a minimum of 4 anymore. It can be tuned via system property reactor.schedulers.defaultPoolSize now though (#1243, #1246)repeat(N) is now aligned with retry(N), as well as repeat(N, Predicate)/retry(N, Predicate): the number indicates the amount of repetitions of the original sequence, ie 0 mirrors the original sequence while 1 mirrors the original sequence + an additional repetition of it. Negative values are rejected at assembly time. (#1252)
repeat(long) in your code and decrement the passed parameter by onerepeat(0) should be replaced with Flux/Mono.empty()windowUntil and windowWhile don't have a final empty window at the end of the sequence (#1033)Operators.onErrorDropped is now to both bubble and throw the error
Mono.fromCompletionStage now drops fatal exceptions rather than hanging (#1118)blockLast(), block(), iterator()) called inside a parallel or single Scheduler trigger an exception (#1102)
concatMapDelayError default behavior is now to delay the errors until the END (#1101)
StepVerifier.withVirtualTime now use a global lock, making virtual time verifications mutually exclusive (#648)
create uses Operators.onOperatorError hook instead of onErrorDropped (#1078)Scannable#operatorName() is deprecated in favor of stepName() (#1115, #1140)Mono.publish(Function) (#437)metrics() to Flux (#1183, #1123, 34e0d4b5, #1245, #1242, #1250)
Micrometer is not on the classpathretryWithBackoff (#1122)
retry(n), the complex retryWhen(Function) and the more configurable RetryFunction from reactor-addonsusingWhen (#1220, b87ea6d, #1233, #1259)
using, but the resource is provided asynchronously through a PublisherFunction<Resource, Publisher>) and only delay the propagation of the terminal signals, NOT the onNext signals.Publisher<Resource> itself is cancelled.doOnDiscard(Class, Consumer) operator (#999, #1343, #1345, #1347)Context hook that will let upstream operators in the chain clean up elements that get either filtered out of the downstream sequence or prefetched then discardedonErrorContinue() mode that influences the upstream chain of operator (#629)
Throwable is passed to a handler instead of downstream.String.replaceAll with a static Pattern and Matcher.replaceAll in Scannable (#1315)limitRate(n, 0) (#1317)
lowTide and highTide parameter could still be changed by internal prefetching strategieslowTide of 0, the requests will strictly adhere to highTidetoStream() and toIterable(), and now only reject when the iteration is performed (#1313)
Iterable was materialized, but what actually blocks is the act of iterating.Mono.from(Flux.from(mono)) is now immediately returning the mono instance, as this is conceptually a no-op. (#1329)
Flux.from(Mono.from(flux))) is NOT a no-op as it transforms the original Flux into a Mono, so it still returns a different instance.Scheduler pool size is now configurable via a system property (#1243)QueueSubscription is logged via log() during an onNext, its toString method is explicitly called (#1270)
window() immediately followed by log(), which is not really useful but was even less useful since it would trigger an exceptionTuple2..Tuple8 now have mapT1..mapT8 (as relevant) methods that allow to change a single part of the Tuple (e4a9aee4, discussed in #1058)cache(Duration) now have an overload that takes a Scheduler, added Mono.cache TTL-generator variant (#1189, #1125)Operators.lift now has an alternative that exposes the raw Publisher: liftPublisher and supports Fuseable (#1205, #1206)zip variants with 7 and 8 arguments (#1210Flux.interval (#1178)empty() queue Supplier to Queues, support empty and one cases in Queues.capacity() (#1161, d9d76aba)doOnEach now support Fuseable and ConditionalSubscriber (#1003)Flux.take(0) now eagerly cancels on subscription (#1158)Mono#fromFuture and Mono#fromCompletionStage added, with Supplier param (#1131)error(Supplier) variant to Flux and Mono (#1100)Scannable#steps(), which produces a Stream<String> of all the stepNames both upstream and downstream (including the current Scannable) (#1115, #1140)Scannable (this avoids problems for users that use instanceof checks in hooks, #1136)Flux.split extension to convert Flux<List<T>> to Flux<T> (#1089)verifyThenAssertThat(Duration) to allow assertions and verification timeout (#1124)LoggerFactory with a custom implementation, provide a TestLogger one that can be used to assert log contents (8f3a8fab)TestPublisher than can be set up before eg. subscribing in a StepVerifier (#1236)doOnSuccess and subscribing without an error handler, errors would be swallowed (#1337)
ErrorCallbackNotImplemented exception like in other cases, unless a doOnTerminate is also usedsubscribe(...)ArrayIndexOutOfBoundsException when using lift with a ParallelFlux operator that is fuseable (can be triggered by using Sleuth, #1293)THREAD_BARRIER in concatMap* and publish (#1302)TestPublisher, calling .mono() would turn it into a compliant one. This has been fixed and the mono will now continue to be a bad boy (#1244)bufferTimeout and bufferWhen issues have been corrected around cancellation and drain race conditions (#1247)Flux.last() now correctly throws a NoSuchElementException on empty source even when said source is a Callable Mono(#1226)delayUntil correctly request Long.MAX_VALUE and not Integer.MAX_VALUE on subscribe (#1214)Operators.lift() properly maintain GroupedFlux/ConnectableFlux interfaces (#1204)Context resolution resolves to 1st context (#1114)WorkQueueProcessor.forceShutdown (#1142)VirtualTimeScheduler defer advanceTime if queue is empty (#783)
subscribeOn, the operators delaying values end up reading the clock AFTER the virtual time has been moved forward by thenAwait, hanging the test.FluxPublish dispose to clear the existing connection (#1173)WorkerTask Future if cancelled race (#1107)FluxRefCount DisconnectException on cancellation race (#1088)FluxGroupBy.checkTerminated (#1094)Mono.fromCompletionStage now drops fatal exceptions rather than hanging (#1118)SpscArrayQueue (35519ab)1.2.51 (#1274)toList() vs iterator() (#1135)doOnSubscribe usage (#1090)A notable change in the way we work with the repository and contributions that make sense for maintenance releases: as of commit 0d8e16e5, there will be no more port #xxx backport commits in maintenance branches, but rather issues will be fixed first in the maintenance branch THEN a forward-merge will happen to include the fix in the master branch. See #1225.
@Alex-Diez, @baptistemesta, @charlesmuchene, @genie-youn, @igoperikov, @madgnome, @MarkusJais, @nurkiewicz, @OlegDokuka, @pdudkiewicz, @smiklos, @szpak, @utwyko, @xcorail, @yamkazy, @0xflotus
Reactor-Core 3.1.16.RELEASE is part of `Bismuth-SR17` Release Train.
Reactor-Core 3.1.16.RELEASE is part of Bismuth-SR17 Release Train.
This is a maintenance update recommended for all Reactor 3 users.
FluxBufferTimeout on flush (#1534)This is a short maintenance release, part of the Bismuth-SR16 Release Train.
This is a short maintenance release, part of the Bismuth-SR16 Release Train.
mergeSequential's first inner sequence would not be cancelled with the rest when output Flux was cancelled (#1500)Mono.fromRunnable can now withhold executing the Runnable if the Subscription is immediately cancelled (#1503)This is the 15th release of Reactor 3.1, part of `BISMUTH-SR15` Release Train.
This is the 15th release of Reactor 3.1, part of BISMUTH-SR15 Release Train.
This release contains minor new API in the Context class.
putNonNull and size methods to Context (#1442)It has been built using JDK11 and the javadoc jars are non-functional, especially when served from the reactor website.
v3.1.14.RELEASE-It has been built using JDK11 and the javadoc jars are non-functional, especially when served from the reactor website.
See release notes at https://github.com/reactor/reactor-core/releases/tag/v3.1.14.RELEASE
This is the 13th release of Reactor 3.1, part of `Bismuth-SR14` Release Train.
This is the 13th release of Reactor 3.1, part of Bismuth-SR14 Release Train.
This is a recommended update for all Reactor 3.1 users.
expectAccessibleContext work with initialContext from StepVerifier (#1417)StepVerifier now ties back to the Context provided in the options.@OlegDokuka
This is the 12th release of Reactor 3.1, part of `Bismuth-SR13` Release Train.
This is the 12th release of Reactor 3.1, part of Bismuth-SR13 Release Train.
This maintenance release contains a few bugfixes.
share() and .publish().refCount() (including one with grace period) missing repeat/retry re-subscribes (#1385, #1386)reduce() lambda sometimes being passed a null value, potentially causing NPE (#1393)This is the 13th release of Reactor 3.1, part of `Bismuth-SR12` Release Train.
This is the 13th release of Reactor 3.1, part of Bismuth-SR12 Release Train.
This release is a small bugfix release, recommended for all 3.1.x users.
EmitterProcessor SYNC fusion completing too early, not producing an element when fused (#1364)CombineLatest ClassCastException with single source(#1356)This is the 10th release of Reactor 3.1, part of `BISMUTH-SR11` Release Train.
This is the 10th release of Reactor 3.1, part of BISMUTH-SR11 Release Train.
This is a recommended update for all Reactor 3.1 users.
QueueSubscription objects as collections in log()'s onNext (#1270)toStream() and toIterable(), and now only reject when the iteration is performed (#1313)
Iterable was materialized, but what actually blocks is the act of iterating.Mono.from(Flux.from(mono)) is now immediately returning the mono instance, as this is conceptually a no-op. (#1329)
Flux.from(Mono.from(flux))) is NOT a no-op as it transforms the original Flux into a Mono, so it still returns a different instance.bufferTimeout and bufferWhen issues (#1247)refCount quick subscribe-and-cancel (#1260)Flux#publish and EmitterProcessor should trigger an extra poll during sync fusion (#1290)ParallelFlux, when decorated through lift() (eg. by instrumenting using Sleuth), would throw an IndexOutOfBoundsException (#1293)Flux#publish, concatMap* could run on the wrong Scheduler when fused (#1302)doOnSuccess and subscribing without an error handler, errors would be swallowed (#1337)
ErrorCallbackNotImplemented exception like in other cases, unless a doOnTerminate is also usedsubscribe(...)create (#1266)This is the 9th release of Reactor 3.1, part of `BISMUTH-SR10` Release Train.
This is the 9th release of Reactor 3.1, part of BISMUTH-SR10 Release Train.
This is a recommended update for all Reactor 3 users. Note that all fixes in this release are backports of fixes in the upcoming 3.2.0.RELEASE.
Flux.last() used to skip throwing a NoSuchElementException on some category of empty sources (Flux or Mono that are Callable). This is a bug and it correctly does throw now. (#1226)Scannable (#1136)Operators.lift function now supports Fuseable (#1206)Flux.last() throws NoSuchElementException (see update considerations section, #1226)Operators.lift properly maintain GroupedFlux/ConnectableFlux interface (#1204)FluxOnBackpressureBufferTimeout (#1194)Flux.interval support 0 delay/period (#1178)delayUntil now requests Long.MAX_VALUE instead of Integer.MAX_VALUE (#1214)@madgnome, @OlegDokuka
This is the 8th release of Reactor 3.1, part of `BISMUTH-SR9` Release Train.
This is the 8th release of Reactor 3.1, part of BISMUTH-SR9 Release Train.
This maintenance update is mainly backporting QoL enhancements and fixes from the ongoing 3.2.0 development effort.
MonoProcessor internals have been improved and now support Context resolution (https://github.com/reactor/reactor-core/commit/6ab518681f42b4e3b683d4d1200b048a5c450784)Flux.take(0) now eagerly cancel and onComplete (#1158)Mono.cache(Duration) should not onSubscribe after adding a new subscriber (#1190)WorkQueueProcessor.forceShutdown should interrupt hanging threads (#1142)Flux.publish should properly dispose its state on cancel (#1173)Many thanks for you reports folks !!! 💃 @vulcannis @rwinch @elrodro83 @stavalfi
Special thanks to @OlegDokuka for his continuous support on our Gitter channel 🥇
This is the 7th release of Reactor 3.1, part of `BISMUTH-SR8` Release Train.
This is the 7th release of Reactor 3.1, part of BISMUTH-SR8 Release Train.
This maintenance update is mainly backporting QoL enhancements and fixes from the ongoing 3.2.0 development effort.
Operators.onErrorDropped still bubbles the exception up, but it also logs the error (#1118, 97d110f3)
Mono.fromFuture change belowparallel() and single() Schedulers (#1102 , 52912391)
Mono#block, Flux#blockFirst, Flux#blockLast, toIterator, toStreamStepVerifier.withVirtualTime is now mutually exclusive to better protect from its side effects on Schedulers when running multiple tests in parallel (#648, 65e7d625)Mono.fromFuture would sometimes swallow errors.
StackOverflowError are bubbled up and are not expected to be part of the sequence anymore, here such error would totally disappear. Now MonoCompletionStage drops such errors, which results in the SOE being logged thanks to the associated change to Operators.onErrorDropped (97d110f3, #1118)WorkerTask Future in case of cancel race (#1107, e8f78678)This is the 5th release of Reactor 3.1, part of `BISMUTH-SR7` Release Train.
This is the 5th release of Reactor 3.1, part of BISMUTH-SR7 Release Train.
This is a recommended update for all Reactor 3 users. Note that all fixes in this release are backports of fixes in the upcoming 3.2.0.RELEASE.
create uses onOperatorError hook instead of onErrorDropped, avoiding thrown exceptions (#1078)RejectedExecutionException now only terminate the ExecutorScheduler if the underlying Executor is shut down (#1080)DisconnectException on cancellation race in ParallelFlux#refCount (#1088)doOnSubscribe javadoc: not for acting on the Subscription (#1090)blockLast javadoc mentioning "first" element instead of "last" (#1095)groupBy code (#1094)@baptistemesta, @smiklos
One column per quarter.