td-agent fluentd无法向Kafka发送消息

Question

我让td-agent将消息发送到Kafka队列，没有任何问题。但是现在我得到了以下错误。我重新启动了代理，但没有成功。我手动刷新了缓冲区，但一无所获。

欢迎任何关于这方面的建议。

2019-03-20 16:44:06 +0100 warn：#0发送异常发生:无法向test发送消息-250-events/0 2019-03-20 16:44:06 +0100 warn：#0异常回溯: /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:330:in deliver_messages_with_retries' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:238:inblock in deliver_messages‘/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/instrumenter.rb:23:in /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1. instrument' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:231:indeliver_messages‘/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.instrument' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:231:indeliver_messages’/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.instrument' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:231:indeliver_messages‘/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.instrument' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:231:indeliver_messages’/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.instrument' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:231:indeliver_messages‘/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.instrument' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/ruby-kafka-0.6.8/lib/kafka/producer.rb:231:indeliver_messages’deliver_messages' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluent-plugin-kafka-0.7.9/lib/fluent/plugin/out_kafka_buffered.rb:344:inwrite‘2.6/lib/fluent/compat/output.rb:131:在start‘/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.2.6/lib/fluent/plugin_helper/thread.中的write' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.2.6/lib/fluent/plugin/output.rb:1111:intry_flush’/opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.2.6/lib/fluent/plugin/output.rb:1394:in flush_thread_run' /opt/td-agent/embedded/lib/ruby/gems/2.4.0/gems/fluentd-1.2.6/lib/fluent/plugin/output.rb:445:inblock (2 levels)中rb:78:in ` `block in thread_create‘2019-03-20 16:44:06 +0100信息：#0初始化kafka生产者: kafka 2019-03-20 16:44:06 +0100警告：#0刷新缓冲区失败。retry_time=5 next_retry_seconds=2019-03-20 16:44:05 +0100 chunk="58471660306e5c7712d0c2ff1a546048“error_class=Kafka::DeliveryFailed error=”无法向test发送消息-250-events/0“2019-03-20 16:44:06 +0100 warn：#0禁止相同的堆栈跟踪

从服务器端连接Kafka集群：

用户@服务器：~$ telnet digital.servicebus.windows.net 9093正在尝试13.69.64.0...已连接到digital-dev-01.servicebus.windows.net。转义字符是'^]‘。

提前谢谢你

Answer 1

我们经历了完全相同的问题。在Azure北欧的Kubernetes集群中使用Kafka输出运行fluentd。

3月19日下午1点左右，它在多个集群上停止工作。

编辑:我们找到了一个解决方案。基本上，请确保您有更高版本的ruby-kafka和/或fluent-kafka-plugin

我们通过在Kubernetes清单中实现了这一点。

容器：- name: fluentd image: fluent/fluentd-kubernetes-daemonset:v1.3.3-debian-kafka-1.2命令："/bin/sh“args："-c"，"cp /fluent-config/*.conf /fluentd/etc/ && gem install fluent-plugin-detect-exceptions && gem install fluent-plugin-prometheus && gem update fluent-plugin-kafka &&/fluentd/entrypoint.sh”