topic Re: Splunk light weight forwarder failover capability in Getting Data In

Splunk light weight forwarder failover capability

ajaybguthi — Fri, 10 May 2013 17:46:46 GMT

Hi,

Do we have a fail over capability for any Splunk forwarders? like if one forwarder goes down the other one will pickup and start reading the files where the other forwarder stopped

Re: Splunk light weight forwarder failover capability

bosburn_splunk — Fri, 10 May 2013 18:06:43 GMT

This functionality is not included in the base Splunk forwarder install. You could use deployment monitor app to watch for forwarders that stop sending data, or write a watchdog script that will keep an eye out on the Splunk process and restart it if it crashes.

Re: Splunk light weight forwarder failover capability

alacercogitatus — Fri, 10 May 2013 18:08:23 GMT

first upgoat!

Re: Splunk light weight forwarder failover capability

dwaddle — Fri, 10 May 2013 18:16:54 GMT

Multiple forwarders on the same machine reading and forwarding the same files is a disaster waiting to happen. Each forwarder keeps a database of what files it has seen and how far into them it has forwarded. A "fail over" forwarder running on the same machine would very likely result in lots of duplicate events.

I would try to think of the Splunk forwarder as an agent. You don't run "fail over" antivirus software, or "fail over" ssh daemons. As bosburn suggests, watching inactive forwarders via deployment monitor or a watchdog to restart a forwarder if it does crash makes sense. But don't overcomplicate it.

Re: Splunk light weight forwarder failover capability

piebob — Fri, 10 May 2013 18:57:19 GMT

inb4--oh dammit.