Fluentd 日志处理-S3拉取日志处理（二）

标签：get   tar   ons   col   xxxxx   保持数据   href   字段   access   

<source>
    @type s3
    aws_key_id       XXXXXXXXXXXXXXXXXXXX
    aws_sec_key    XXXXXXXXXXXXXXXXXXXX
    s3_bucket        nx-rc-rancher-newelb-inner
    s3_region         cn-northwest-1
    tagelb-inner
    <sqs>
       queue_name  fluentd-s3      #这里只需要配置SQS的名字
    </sqs>
</source>
<source>
    @type s3
    aws_key_id      XXXXXXXXXXXXXXXXXXXX
    aws_sec_key    XXXXXXXXXXXXXXXXXXXX
    s3_bucket       nx-rc-rancher-newelb-com
    s3_region       cn-northwest-1
    <sqs>
       queue_name  fluentd-s3
    </sqs>
</source>
<source>
    @type s3
    aws_key_id       XXXXXXXXXXXXXXXXXXXX
    aws_sec_key    XXXXXXXXXXXXXXXXXXXX
    s3_bucket        nx-rc-www-newelb-com
    s3_region        cn-northwest-1
    <sqs>
queue_name      fluentd-s3
    </sqs>
</source>

这里的rewrite_tag_filter插件相当于logstash的if语句。

因为fluentd是通过流的形式进行信息的过滤和处理，而且没有if语句，只能通过重写tag来筛选不同的日志
官方文档
https://github.com/fluent/fluent-plugin-rewrite-tag-filter

<match elb-inner>
    @type rewrite_tag_filter
    <rule>
        key     message
        pattern /.yufuid.net:80/
        tag     elb-inner.net      #匹配的就重写标签
    </rule>
</match>
<filter elb-inner.net>
    @type parser                   #对日志就行分割和命名处理，fluentd的正则表达式可能需要自己写
    key_name message
    reserve_data yes
    <parse>
    @type regexp  
    expression /(?<elb_http_method>[^ ]+) (?<access_timestamp>[^ ]+) (?<elb_name>[^ ]+) (?<client_ip>[^ ]+):(?<client_Port>[^ ]+) (?<target_ip_port>[^ ]+) (?<request_processing_time>[^ ]+) (?<target_processing_time>[^ ]+) (?<response_processing_time>[^ ]+) (?<elb_status_code>[^ ]+) (?<target_status_code>[^ ]+) (?<received_bytes>[^ ]+) (?<send_bytes>[^ ]+) (?<request>"[^\"]+") (?<client_info>"[^"]+") (?<ssl_cipher>[^ ]+) (?<ssl_protocol>[^ ]+) (?<target_group_arn>[^ ]+) (?<trace_ip>[^ ]+) (?<domainname>[^ ]+) (?<chose_cert_arn>[^ ]+) (?<matched_rule_priority>[^ ]+) (?<elb_name>[^ ]+) (?<request_creation_time>[^ ]+)/
    </parse>
</filter>
<filter elb-com>
    @type parser
    key_name message
    reserve_data yes
    <parse>
    @type regexp
            expression /(?<elb_http_method>[^ ]+) (?<access_timestamp>[^ ]+) (?<elb_name>[^ ]+) (?<client_ip>[^ ]+):(?<client_Port>[^ ]+) (?<target_ip_port>[^ ]+) (?<request_processing_time>[^ ]+) (?<target_processing_time>[^ ]+) (?<response_processing_time>[^ ]+) (?<elb_status_code>[^ ]+) (?<target_status_code>[^ ]+) (?<received_bytes>[^ ]+) (?<send_bytes>[^ ]+) (?<request>"[^\"]+") (?<client_info>"[^"]+") (?<ssl_cipher>[^ ]+) (?<ssl_protocol>[^ ]+) (?<target_group_arn>[^ ]+) (?<trace_ip>[^ ]+) (?<domainname>[^ ]+) (?<chose_cert_arn>[^ ]+) (?<matched_rule_priority>[^ ]+) (?<elb_name>[^ ]+) (?<request_creation_time>[^ ]+)/
    </parse>
</filter>
<filter elb-www>
    @type parser
    key_name message
    reserve_data yes
    <parse>
        @type regexp
        expression  /(?<elb_http_method>[^ ]+) (?<access_timestamp>[^ ]+) (?<elb_name>[^ ]+) (?<client_ip>[^ ]+):(?<client_Port>[^ ]+) (?<target_ip_port>[^ ]+) (?<request_processing_time>[^ ]+) (?<target_processing_time>[^ ]+) (?<response_processing_time>[^ ]+) (?<elb_status_code>[^ ]+) (?<target_status_code>[^ ]+) (?<received_bytes>[^ ]+) (?<send_bytes>[^ ]+) (?<request>"[^\"]+") (?<client_info>"[^"]+") (?<ssl_cipher>[^ ]+) (?<ssl_protocol>[^ ]+) (?<target_group_arn>[^ ]+) (?<trace_ip>[^ ]+) (?<domainname>[^ ]+) (?<chose_cert_arn>[^ ]+) (?<matched_rule_priority>[^ ]+) (?<elb_name>[^ ]+) (?<request_creation_time>[^ ]+)/
    </parse>
</filter>

这里是把处理过的日志数据输出到ES集群

<match elb-www>
    @type elasticsearch
    host elasticsearchlog-lb.elasticsearch-log
    index_name    fluentd-www-elb
    type_name     fluentd-www-elb
    flush_interval 2s                     #这里的刷新是把buffer的数据及时发送到ES，保持数据的实时性
    include_timestamp true         #这里需配置时间字段，便于kibana的时间跟踪
    ssl_verify    false
</match>
<match elb-com>
    @type elasticsearch
    host elasticsearchlog-lb.elasticsearch-log
    index_name    fluentd-elb
    type_name     fluentd-elb
    flush_interval 2s
    include_timestamp true
    ssl_verify    false
</match>
<match elb-inner.net>
    @type elasticsearch
    host elasticsearchlog-lb.elasticsearch-log
    index_name    fluentd-elb
    type_name     fluentd-elb
    flush_interval 2s
    include_timestamp true
    ssl_verify    false
</match>

Fluentd 日志处理-S3拉取日志处理（二）

标签：get   tar   ons   col   xxxxx   保持数据   href   字段   access   

原文地址：http://blog.51cto.com/11078047/2316910