Filebeat日志采集使用方式
最近由于业务需求,需要通过filebeat采集日志文件到kafka中,供flink消费,其具体步骤如下:
-
安装filebeat
-
文件配置
-
filebeat-input.yml文件配置
- backoff: 1s backoff_factor: 2 clean_inactive: 25h close_inactive: 2h encoding: plain fields: {NODE_IP: '日志所在节点', target_topic: 目标kafka topic} fields_under_root: true force_close_files: false harvester_buffer_size: 16384 ignore_older: 24h max_backoff: 10s max_bytes: 10485760 multiline: {match: after, negate: true, pattern: '^\d{4}\-\d{1,2}\-\d{1,2}\s\d{2}:\d{2}:\d{2}.\d{3}\s\001'} paths: [日志所在文件] scan_frequency: 1s tail_files: true type: log
-
filebeat.yml文件配置
queue: mem: events: 2048 flush.min_events: 1024 # 日志文件采集配置 filebeat.config.inputs: enabled: true path: conf/filebeat-input.yml reload.enabled: true reload.period: 10s # 输出到终端,供调试用 #output.console: # pretty: ${CONSOLE:false} # 输出到kafka output.kafka: enabled: true hosts: ["kafka所在节点"] topic: '%{[target_topic]}' partition.round_robin: reachable_only: false required_acks: 1 compression: gzip max_message_bytes: 1000000
-
-
配置好上述文件以后,就可以启动filebeat进行日志采集:
./filebeat -e -c /app/filebeat/conf/filebeat.yml
-
在kafka测试,即可收到对应日志