Kafka 服务隔几天就卡死一次 kill-9 之后 重启又能恢复 麻烦各位大佬看看是啥原因

[Arrackisarookie]

Kafka 服务隔几天就卡死一次 kill-9 之后 重启能恢复正常 过几天又不行了

麻烦各位大佬看看是啥原因，看看是性能问题还是网络问题还是啥别的

版本 kafka3.0.0 ，用的自带的 zookeeper ，数据量不小，每个主题保存了 5G ，大概最多几个小时的数据

用了自带的 sasl-plaintext 鉴权，单服务器单节点

broker.id=0
listeners=LOCAL://xxx:9092,JTKG://xxx:9093,GDKJ://xxx:9094  # 跨网所以映射了多个，这里的 xxx 代表同一个 ip
advertised.listeners=LOCAL://xxx:9092,JTKG://yyy:9093,GDKJ://zzz:29092  # xxx,yyy,zzz 是不同 ip
listener.security.protocol.map=LOCAL:SASL_PLAINTEXT,JTKG:SASL_PLAINTEXT,GDKJ:SASL_PLAINTEXT
inter.broker.listener.name=LOCAL
sasl.enabled.mechanisms=SCRAM-SHA-256
sasl.mechanism.inter.broker.protocol=SCRAM-SHA-256
authorizer.class.name=kafka.security.authorizer.AclAuthorizer
super.users=User:admin
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1024000
socket.receive.buffer.bytes=1024000
socket.request.max.bytes=104857600
log.dirs=/home/kafka/kafka_2.12-3.0.0/kafka-logs
num.partitions=1
num.recovery.threads.per.data.dir=1
offsets.topic.replication.factor=1
transaction.state.log.replication.factor=1
transaction.state.log.min.isr=1
log.retention.hours=168
log.retention.bytes=5368709120
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
zookeeper.connect=localhost:2181
zookeeper.connection.timeout.ms=60000
group.initial.rebalance.delay.ms=0

下为最近一次日志

[2024-12-19 13:14:31,683] INFO [Controller id=0] Partitions undergoing preferred replica election:  (kafka.controller.KafkaController)
[2024-12-19 13:14:31,683] INFO [Controller id=0] Partitions that completed preferred replica election:  (kafka.controller.KafkaController)
[2024-12-19 13:14:31,692] INFO [Controller id=0] Skipping preferred replica election for partitions due to topic deletion:  (kafka.controller.KafkaController)
[2024-12-19 13:14:31,692] INFO [Controller id=0] Resuming preferred replica election for partitions:  (kafka.controller.KafkaController)
[2024-12-19 13:14:31,693] INFO [Controller id=0] Starting replica leader election (PREFERRED) for partitions  triggered by ZkTriggered (kafka.controller.KafkaController)
[2024-12-19 13:14:31,823] INFO [Controller id=0] Starting the controller scheduler (kafka.controller.KafkaController)
[2024-12-19 13:14:31,847] DEBUG [Controller id=0] Resigning (kafka.controller.KafkaController)
[2024-12-19 13:14:31,847] DEBUG [Controller id=0] Unregister BrokerModifications handler for Set(0) (kafka.controller.KafkaController)
[2024-12-19 13:14:31,859] INFO [PartitionStateMachine controllerId=0] Stopped partition state machine (kafka.controller.ZkPartitionStateMachine)
[2024-12-19 13:14:31,859] INFO [ReplicaStateMachine controllerId=0] Stopped replica state machine (kafka.controller.ZkReplicaStateMachine)
[2024-12-19 13:14:32,002] INFO [RequestSendThread controllerId=0] Shutting down (kafka.controller.RequestSendThread)
[2024-12-19 13:14:32,051] WARN [RequestSendThread controllerId=0] Controller 0 epoch 60 fails to send request (xxx) to broker xxx:9092 (id: 0 rack: null). Reconnecting to broker. (kafka.controller.RequestSendThread)
java.io.IOException: Client was shutdown before response was read
    at org.apache.kafka.clients.NetworkClientUtils.sendAndReceive(NetworkClientUtils.java:109)
    at kafka.controller.RequestSendThread.doWork(ControllerChannelManager.scala:252)
    at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:96)
[2024-12-19 13:14:32,070] ERROR [RequestSendThread controllerId=0] Controller 0 fails to send a request to broker xxx:9092 (id: 0 rack: null) (kafka.controller.RequestSendThread)
java.lang.InterruptedException
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.tryAcquireSharedNanos(AbstractQueuedSynchronizer.java:1326)
    at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:277)
    at kafka.utils.ShutdownableThread.pause(ShutdownableThread.scala:82)
    at kafka.controller.RequestSendThread.backoff$1(ControllerChannelManager.scala:233)
    at kafka.controller.RequestSendThread.doWork(ControllerChannelManager.scala:261)
    at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:96)
[2024-12-19 13:14:32,071] INFO [RequestSendThread controllerId=0] Stopped (kafka.controller.RequestSendThread)
[2024-12-19 13:14:32,076] INFO [RequestSendThread controllerId=0] Shutdown completed (kafka.controller.RequestSendThread)
[2024-12-19 13:14:32,146] INFO [Controller id=0] Resigned (kafka.controller.KafkaController)
[2024-12-19 13:14:32,316] DEBUG [Controller id=0] Broker 0 has been elected as the controller, so stopping the election process. (kafka.controller.KafkaController)
[2024-12-19 13:14:51,582] INFO [ControllerEventThread controllerId=0] Shutting down (kafka.controller.ControllerEventManager$ControllerEventThread)
[2024-12-19 13:14:51,582] INFO [ControllerEventThread controllerId=0] Stopped (kafka.controller.ControllerEventManager$ControllerEventThread)
[2024-12-19 13:14:51,582] INFO [ControllerEventThread controllerId=0] Shutdown completed (kafka.controller.ControllerEventManager$ControllerEventThread)
[2024-12-19 13:14:51,583] DEBUG [Controller id=0] Resigning (kafka.controller.KafkaController)
[2024-12-19 13:14:51,583] DEBUG [Controller id=0] Unregister BrokerModifications handler for Set() (kafka.controller.KafkaController)
[2024-12-19 13:14:51,583] INFO [PartitionStateMachine controllerId=0] Stopped partition state machine (kafka.controller.ZkPartitionStateMachine)
[2024-12-19 13:14:51,583] INFO [ReplicaStateMachine controllerId=0] Stopped replica state machine (kafka.controller.ZkReplicaStateMachine)
[2024-12-19 13:14:51,583] INFO [Controller id=0] Resigned (kafka.controller.KafkaController)

[yoruoxx]

我觉得可以问问 Claude 😂

[fox2081]

居然能撞头像的🙃

[JavasBoy]

系统多大内存，分配了多少 JVM,这些列出来，系统负载监控，包括网卡流量，CPU 、IO 、IOPS 。系统日志。

[rekulas]

以我对 java 的固有印象 内存冗余多吗

[Arrackisarookie]

@yoruoxx 免费版的不太好使😣

@fox2081 hhhh 这不是巧了么

@JavasBoy
@rekulas
之前用的都是默认的 Kafka 4G ，zookeeper 512M ，昨天把 Kafka JVM 的-Xmx 加到了 20G ，zookeeper 的加到了 4G ，这台服务器上就这俩服务，CPU 基本没超过 40%过。
所以大佬，一般这种现象基本都是资源不足的问题是么？应该不是服务本身配置项的问题吧？我问通义它提到了可能需要调整 `log.flush.interval.messages` 和 `log.flush.interval.ms`啥的