Logstash: getting hadoop's tasks-logs

(OhadR) #1

I've searched here and googled the web, but surprisingly I could not find any info about this subject.
So I use LS+ES+Kibana to collect and view logs of services that run on tomcat. Works like a charm.
Now my next challenge is to collect logs of hadoop tasks. I saw this plugin, but the author told me it deals only with hadoop deamons, rather than logs of tasks...

hadoop tasks' logs are log4j, so if I use log4j-input-plugin, and use SocketAppender in my log4j.xml, I do get the logs, but I cannot know from which TASK it comes from (there can be several tasks on the same host)

anyone has an idea? I'm sure I'm not the only one who tries to see hadoop logs with logstash....

(system) #2