I was referring to this link, https://wiki.splunk.com/Community:Best_Practice_For_Configuring_Syslog_Input to configure data buffering on Heavy Forwarder. The instructions provided talk about configuring the below attributes in outputs.conf,
However, except maxQueueSize, no other attributes can be found in the outputs.conf file.
Please advise how can I configure data buffering in this case.
1 is in outputs.conf
and 2 and 4 are deprecated. Unfortunately, the community wiki is pretty out of date and not really supported. For example, it suggests Heavy Forwarder which is actually VERY wrong in the latest versions (should always be UF):
https://www.splunk.com/blog/2016/12/12/universal-or-heavy-that-is-the-question.html
So you mean UF can be used for data buffering ? if yes, can you please share the documentation for the configuration ? Thanks
No, HF is supported but only recommended if:
1: You have to transform most of the data before sending it to indexers
OR
2: You are using Modular Inputs
Otherwise, use UF as the aggregation.
Hi @woodcock, sorry if my question isn't clear enough.
What my main question is , how I can configure Heavy Forwarder to buffer data if Indexer goes down ?
You have mentioned that the 1, 2 and 4 settings are deprecated, but HF still supports data buffering.
So, can you please suggest what are the other steps required to configure that ?
Your best bet is to read these 4 files:
$SPLUNK_HOME/etc/system/README/outputs.conf.spec
$SPLUNK_HOME/etc/system/README/outputs.conf.examples
$SPLUNK_HOME/etc/system/README/limits.conf.spec
$SPLUNK_HOME/etc/system/README/limits.conf.examples