Closed Aloren closed 4 years ago
We will integrate these changes in the next java client release.
This pull request conflicts with a previous pull request (https://github.com/aerospike/aerospike-client-java/pull/152). Can you resolve the conflicts with the previous pull request and resubmit?
@BrianNichols Not sure how do you expect me to resolve the conflicts if none of the branches is merged? It would be much simpler to rebase the second pull request to the master if any of the pull requests gets merged.
This repo has been updated with the previous merged pull request (https://github.com/aerospike/aerospike-client-java/pull/152). You can now modify this pull request to fix the merge conflicts.
@BrianNichols I've updated the branch, thank you
This pull request has been accepted.
reactor-client tests are failing. Not sure which pull request caused the failure, but this needs to be fixed before java client release.
Running com.aerospike.client.reactor.BatchReactorFailTest
SLF4J: No SLF4J providers were found.
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#noProviders for further details.
Tests run: 12, Failures: 4, Errors: 0, Skipped: 0, Time elapsed: 1.678 sec <<< FAILURE! - in com.aerospike.client.reactor.BatchReactorFailTest
shouldFailOnBatchGetSequence[eventLoopType=DIRECT_NIO](com.aerospike.client.reactor.BatchReactorFailTest) Time elapsed: 0.329 sec <<< FAILURE!
java.lang.AssertionError: expectation "expectError(Class)" failed (expected: onError(Timeout); actual: onNext(test:test:batchkey3:49fb2914d0d642dc30dc0a5421ff0b5b10d4748e:(gen:63),(exp:320793698),(bins:(batchbin:batchvalue3))))
at reactor.test.DefaultStepVerifierBuilder.failPrefix(DefaultStepVerifierBuilder.java:2235)
at reactor.test.DefaultStepVerifierBuilder.fail(DefaultStepVerifierBuilder.java:2231)
at reactor.test.DefaultStepVerifierBuilder.lambda$expectError$6(DefaultStepVerifierBuilder.java:275)
at reactor.test.DefaultStepVerifierBuilder$SignalEvent.test(DefaultStepVerifierBuilder.java:1973)
at reactor.test.DefaultStepVerifierBuilder$DefaultVerifySubscriber.onSignal(DefaultStepVerifierBuilder.java:1326)
at reactor.test.DefaultStepVerifierBuilder$DefaultVerifySubscriber.onExpectation(DefaultStepVerifierBuilder.java:1274)
at reactor.test.DefaultStepVerifierBuilder$DefaultVerifySubscriber.onNext(DefaultStepVerifierBuilder.java:958)
at reactor.core.publisher.FluxCreate$BufferAsyncSink.drain(FluxCreate.java:726)
at reactor.core.publisher.FluxCreate$BufferAsyncSink.next(FluxCreate.java:657)
at reactor.core.publisher.FluxCreate$SerializedSink.next(FluxCreate.java:151)
at com.aerospike.client.reactor.listeners.ReactorRecordSequenceListener.onRecord(ReactorRecordSequenceListener.java:36)
at com.aerospike.client.async.AsyncBatch$GetSequenceCommand.parseRow(AsyncBatch.java:362)
at com.aerospike.client.async.AsyncMultiCommand.parseResult(AsyncMultiCommand.java:114)
at com.aerospike.client.async.AsyncCommand.parseCommandResult(AsyncCommand.java:176)
at com.aerospike.client.async.NioCommand.parseGroupBody(NioCommand.java:569)
at com.aerospike.client.async.NioCommand.readMultiBody(NioCommand.java:508)
at com.aerospike.client.async.NioCommand.readMultiHeader(NioCommand.java:500)
at com.aerospike.client.async.NioCommand.read(NioCommand.java:412)
at com.aerospike.client.async.NioCommand.processEvent(NioCommand.java:276)
at com.aerospike.client.async.NioEventLoop.runCommands(NioEventLoop.java:205)
at com.aerospike.client.async.NioEventLoop.run(NioEventLoop.java:163)
at java.lang.Thread.run(Thread.java:748)
The reactor-client unit tests are still failing on my multi-node cluster. I looked at the incremental source code changes and nothing of significance has changed, so I'm not surprised.
Note that the same reactor-client unit tests work fine on a single-node cluster . Try running the reactor-client unit tests on a 2 node cluster.
@kptfh please take a look
It is possible to create index for the same bin, but with different type; such indexes will prevent QueryIndex from loading indexes cache.