Hi Dave,
Thanks for verifying this release candidate. I have setup the Pulsar
cluster with 3.0 on my laptop and test the produce and consume, it works
fine. We also deploy the it on k8s, the produce, consume and health check
also works fine. Would you please share your broker configurations?
Thanks,
Hang
Dave Fisher <wave4d...@comcast.net> 于2023年4月13日周四 06:49写道：
> Hi -
>
> We do End to end testing of Pulsar versions on GCP and with Pulsar 3.0
> today we have the following problem.
>
> This image shows the broker deployments not coming up. This is a non-TLS
> setup.
>
> [image: Screen Shot 2023-04-12 at 1.09.59 PM.png]
>
> The Broker Pods have errors like this one:
>
> 2023-04-12T20:10:35,995+0000 [broker-topic-workers-OrderedExecutor-3-0] ERROR
> org.apache.pulsar.common.protocol.Commands -
> [PersistentSubscription{topic=persistent://pulsar/pulsar/10.236.0.4:8080/healthcheck,
> name=healthCheck-04a8e22e-3e0b-4aa3-9141-d8970c1b0712}] [-1] Failed to parse
> message metadata
> java.lang.IndexOutOfBoundsException: readerIndex(96) + length(2) exceeds
> writerIndex(96): UnpooledDuplicatedByteBuf(ridx: 96, widx: 96, cap: 96,
> unwrapped: CompositeByteBuf(ridx: 96, widx: 96, cap: 96, components=2))
> at
> io.netty.buffer.AbstractByteBuf.checkReadableBytes0(AbstractByteBuf.java:1442)
> ~[io.netty-netty-buffer-4.1.89.Final.jar:4.1.89.Final]
> at io.netty.buffer.AbstractByteBuf.readShort(AbstractByteBuf.java:749)
> ~[io.netty-netty-buffer-4.1.89.Final.jar:4.1.89.Final]
> at
> org.apache.pulsar.common.protocol.Commands.skipBrokerEntryMetadataIfExist(Commands.java:1692)
> ~[org.apache.pulsar-pulsar-common-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.common.protocol.Commands.parseMessageMetadata(Commands.java:452)
> ~[org.apache.pulsar-pulsar-common-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.common.protocol.Commands.parseMessageMetadata(Commands.java:445)
> ~[org.apache.pulsar-pulsar-common-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.common.protocol.Commands.peekMessageMetadata(Commands.java:1899)
> ~[org.apache.pulsar-pulsar-common-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.common.protocol.Commands.peekAndCopyMessageMetadata(Commands.java:1918)
> ~[org.apache.pulsar-pulsar-common-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.broker.service.AbstractBaseDispatcher.filterEntriesForConsumer(AbstractBaseDispatcher.java:142)
> ~[org.apache.pulsar-pulsar-broker-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.broker.service.AbstractBaseDispatcher.filterEntriesForConsumer(AbstractBaseDispatcher.java:100)
> ~[org.apache.pulsar-pulsar-broker-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.broker.service.persistent.PersistentDispatcherSingleActiveConsumer.internalReadEntriesComplete(PersistentDispatcherSingleActiveConsumer.java:210)
> ~[org.apache.pulsar-pulsar-broker-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.pulsar.broker.service.persistent.PersistentDispatcherSingleActiveConsumer.lambda$readEntriesComplete$1(PersistentDispatcherSingleActiveConsumer.java:151)
> ~[org.apache.pulsar-pulsar-broker-3.0.0-SNAPSHOT.jar:3.0.0-SNAPSHOT]
> at
> org.apache.bookkeeper.common.util.SingleThreadExecutor.safeRunTask(SingleThreadExecutor.java:137)
> ~[org.apache.bookkeeper-bookkeeper-common-4.16.0.jar:4.16.0]
> at
> org.apache.bookkeeper.common.util.SingleThreadExecutor.run(SingleThreadExecutor.java:113)
> ~[org.apache.bookkeeper-bookkeeper-common-4.16.0.jar:4.16.0]
> at
> io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
> ~[io.netty-netty-common-4.1.89.Final.jar:4.1.89.Final]
> at java.lang.Thread.run(Thread.java:833) ~[?:?]
>
> We looked into this but aren’t sure what would cause this to suddenly be
> an issue. The candidate would be differences in the new version of BK …
>
> Best,
> Dave
>
