7

我正在尝试使用以下配置在 fluentd 中跟踪多个日志：

<source>
  type tail
  tag es.workers.worker1

  format /^\[(?<timestamp>.*? .*?) (?<log_level>[INFO|ERROR][^\]]*)\] (?<message>.*)$/

  path /var/log/upstart/worker1.log
  pos_file /var/lib/fluentd/pos/-var-log-upstart-worker1.log.pos

</source>
<source>
  type tail
  tag es.workers.worker2

  format /^\[(?<timestamp>.*? .*?) (?<log_level>[INFO|ERROR][^\]]*)\] (?<message>.*)$/

  path /var/log/upstart/worker2.log
  pos_file /var/lib/fluentd/pos/-var-log-upstart-worker2.log.pos

</source>
<source>
  type tail
  tag es.workers.worker3

  format /^\[(?<timestamp>.*? .*?) (?<log_level>[INFO|ERROR][^\]]*)\] (?<message>.*)$/

  path /var/log/upstart/worker3.log
  pos_file /var/lib/fluentd/pos/-var-log-upstart-worker3.log.pos

</source>
<source>
  type tail
  tag es.workers.worker4

  format /^\[(?<timestamp>.*? .*?) (?<log_level>[INFO|ERROR][^\]]*)\] (?<message>.*)$/

  path /var/log/upstart/worker4.log
  pos_file /var/lib/fluentd/pos/-var-log-upstart-worker4.log.pos

</source>

这是行不通的。通常（但并非总是），我只获取第一个文件的日志。有时它是一个不同的文件，但它总是只有一个。关于发生了什么的任何想法？我在 fluentd 错误日志中没有收到任何有意义的错误。

4

1 回答 1

1

拖尾多个文件可以这样完成（标签将基于文件名）

<source>
  @type tail
  @id in_tail_container_logs
  path /var/lib/docker/containers/*/*-json.log
  pos_file /fluentd/log/containers.log.pos
  time_format "%Y-%m-%dT%H:%M:%S.%L%Z"
  keep_time_key true
  read_from_head true
  tag "docker.*"
  format json
</source>

或者像这样

<source>
  @type tail
  @id in_tail_fos_logs
  @label @LOGS
  path /www/web/log/*.log,/www/web2/log/*.log,/www/web3/log/*.log   
  exclude_path ["/www/web/log/logstash_*.log"]
  pos_file /var/log/td-agent/logs.log.pos
  time_format "%Y-%m-%dT%H:%M:%S.%L%Z"
  read_from_head true
  tag "rowlogs.*"
  format none
</source>
于 2020-02-18T22:34:03.163 回答