cloudfoundry / cloudfoundry/cf-java-client
OutOfMemory/Slow performance upon service creation/deletion
Nobody has claimed this yet.
- Dominant language
- Java
- Stars
- 334
- Forks
- 319
- PR merge metrics
- No merged PRs in 30d
Description
Hi Friends,
we recently noticed a strange behaviour when creating/deleting service instances using cf-java-client that leads to OutOfMemoryException and really slow performance. The issue occurs upon unsuccessful service creation/deletion, when it takes about 20 seconds to build the underlying reactor stacktrace. We confirmed that the stacktrace building is the time consuming part of the execution via heapdumps, threadumps and profiling of the app. The heapdump shows the impressive 150MB char array containing the error message and stacktrace. The truncated stacktrace has the following structure:
reactor.core.publisher.Sinks$EmissionException: Spec. Rule 1.3 - onSubscribe, onNext, onError and onComplete signaled to a Subscriber MUST be signaled serially.
at reactor.core.publisher.InternalManySink.emitNext(InternalManySink.java:56)
Suppressed: reactor.core.publisher.FluxOnAssembly$OnAssemblyException:
Assembly trace from producer [reactor.core.publisher.MonoCacheTime] :
reactor.core.publisher.Mono.checkpoint(Mono.java:1902)
org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:261)
Error has been observed at the following site(s):
|_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:261)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.client.v2.spaces.ReactorSpaces.listServices(ReactorSpaces.java:232)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.operations.services.DefaultServices.createInstance(DefaultServices.java:176)
|_____ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:261)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.client.v2.spaces.ReactorSpaces.listServices(ReactorSpaces.java:232)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.operations.services.DefaultServices.createInstance(DefaultServices.java:176)
|_________ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:261)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.client.v2.spaces.ReactorSpaces.listServices(ReactorSpaces.java:232)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.operations.services.DefaultServices.createInstance(DefaultServices.java:176)
|_____________ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:261)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.client.v2.spaces.ReactorSpaces.listServiceInstances(ReactorSpaces.java:226)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.operations.services.DefaultServices.deleteInstance(DefaultServices.java:215)
|_________________ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:261)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.client.v2.spaces.ReactorSpaces.listServiceInstances(ReactorSpaces.java:226)
|_ Mono.checkpoint ⇢ at org.cloudfoundry.operations.services.DefaultServices.deleteInstance(DefaultServices.java:215)
... <much more data that look identically (a couple of tenths of thousands of lines>
Stack trace:
at reactor.core.publisher.InternalManySink.emitNext(InternalManySink.java:56)
at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.lambda$null$3(AbstractUaaTokenProvider.java:209)
at java.util.Optional.ifPresent(Optional.java:159)
at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.lambda$extractRefreshToken$4(AbstractUaaTokenProvider.java:198)
at reactor.core.publisher.FluxPeek$PeekSubscriber.onNext(FluxPeek.java:184)
at reactor.core.publisher.FluxMap$MapSubscriber.onNext(FluxMap.java:120)
at reactor.core.publisher.FluxMap$MapSubscriber.onNext(FluxMap.java:120)
at reactor.core.publisher.FluxHandle$HandleSubscriber.onNext(FluxHandle.java:118)
at reactor.core.publisher.FluxMap$MapConditionalSubscriber.onNext(FluxMap.java:220)
at reactor.core.publisher.FluxDoFinally$DoFinallySubscriber.onNext(FluxDoFinally.java:130)
at reactor.core.publisher.FluxHandleFuseable$HandleFuseableSubscriber.onNext(FluxHandleFuseable.java:184)
at reactor.core.publisher.FluxContextWrite$ContextWriteSubscriber.onNext(FluxContextWrite.java:107)
at reactor.core.publisher.Operators$MonoSubscriber.complete(Operators.java:1789)
at reactor.core.publisher.MonoCollectList$MonoCollectListSubscriber.onComplete(MonoCollectList.java:128)
at reactor.core.publisher.FluxPeek$PeekSubscriber.onComplete(FluxPeek.java:259)
at reactor.core.publisher.FluxMap$MapSubscriber.onComplete(FluxMap.java:142)
at reactor.core.publisher.FluxDoFinally$DoFinallySubscriber.onComplete(FluxDoFinally.java:145)
at reactor.core.publisher.FluxMap$MapSubscriber.onComplete(FluxMap.java:142)
at reactor.netty.channel.FluxReceive.onInboundComplete(FluxReceive.java:389)
at reactor.netty.channel.ChannelOperations.onInboundComplete(ChannelOperations.java:396)
at reactor.netty.channel.ChannelOperations.terminate(ChannelOperations.java:452)
at reactor.netty.http.client.HttpClientOperations.onInboundNext(HttpClientOperations.java:664)
at reactor.netty.channel.ChannelOperationsHandler.channelRead(ChannelOperationsHandler.java:94)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:379)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:365)
at io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:357)
at io.netty.handler.codec.MessageToMessageDecoder.channelRead(MessageToMessageDecoder.java:103)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:379)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:365)
at io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:357)
at io.netty.channel.CombinedChannelDuplexHandler$DelegatingChannelHandlerContext.fireChannelRead(CombinedChannelDuplexHandler.java:436)
at io.netty.handler.codec.ByteToMessageDecoder.fireChannelRead(ByteToMessageDecoder.java:324)
at io.netty.handler.codec.ByteToMessageDecoder.channelRead(ByteToMessageDecoder.java:296)
at io.netty.channel.CombinedChannelDuplexHandler.channelRead(CombinedChannelDuplexHandler.java:251)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:379)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:365)
at io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:357)
at io.netty.handler.ssl.SslHandler.unwrap(SslHandler.java:1533)
at io.netty.handler.ssl.SslHandler.decodeJdkCompatible(SslHandler.java:1282)
at io.netty.handler.ssl.SslHandler.decode(SslHandler.java:1329)
at io.netty.handler.codec.ByteToMessageDecoder.decodeRemovalReentryProtection(ByteToMessageDecoder.java:508)
at io.netty.handler.codec.ByteToMessageDecoder.callDecode(ByteToMessageDecoder.java:447)
at io.netty.handler.codec.ByteToMessageDecoder.channelRead(ByteToMessageDecoder.java:276)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:379)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:365)
at io.netty.channel.AbstractChannelHandlerContext.fireChannelRead(AbstractChannelHandlerContext.java:357)
at io.netty.channel.DefaultChannelPipeline$HeadContext.channelRead(DefaultChannelPipeline.java:1410)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:379)
at io.netty.channel.AbstractChannelHandlerContext.invokeChannelRead(AbstractChannelHandlerContext.java:365)
at io.netty.channel.DefaultChannelPipeline.fireChannelRead(DefaultChannelPipeline.java:919)
at io.netty.channel.epoll.AbstractEpollStreamChannel$EpollStreamUnsafe.epollInReady(AbstractEpollStreamChannel.java:795)
at io.netty.channel.epoll.EpollEventLoop.processReady(EpollEventLoop.java:480)
at io.netty.channel.epoll.EpollEventLoop.run(EpollEventLoop.java:378)
at io.netty.util.concurrent.SingleThreadEventExecutor$4.run(SingleThreadEventExecutor.java:989)
at io.netty.util.internal.ThreadExecutorMap$2.run(ThreadExecutorMap.java:74)
at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
at java.lang.Thread.run(Thread.java:836)
Suppressed: java.lang.Exception: #block terminated with an error
at reactor.core.publisher.BlockingSingleSubscriber.blockingGet(BlockingSingleSubscriber.java:99)
at reactor.core.publisher.Mono.block(Mono.java:1703)
... <more suppressed exceptions (about 4000) that are exactly the same>
Digging into the cf-java-client/reactor code, we noticed that the reactor provides a functionality for creating a light checkpoint that has no the overhead of stacktrace building upon error, but the wrapping cf-java-client does not expose that. Do you plan to expose such functionality? Or have you observed similar behaviour, any idea how we could overcome this?
We use version 5.1.0 of cf-java-client.
Regards,
Martin
PS. Having a quick overview of the reported issues, I noticed this one - https://github.com/cloudfoundry/cf-java-client/issues/1090 - which might be related to this problem, as well.
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by tracing the unsuccessful service creation and deletion paths through DefaultServices.createInstance/deleteInstance and ReactorSpaces.listServices/listServiceInstances, then inspect AbstractUaaTokenProvider.token and its checkpoint usage. Use the reported cf-java-client 5.1.0 behavior as the baseline; done means the client exposes the requested lightweight checkpoint behavior without the reported stacktrace-building overhead.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- java
- Domain
- backend, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100