Heavy forwarder for Splunk app for AWS

[sreerajms]

In am trying to install and configure Splunk App for AWS and its add-on. We have a non-clustered distributed environment with a search head and three indexers. The documentation at http://docs.splunk.com/Documentation/AWS/5.1.0/Installation/Installon-prem says:

1) Install both the app and add-on to your search heads.
2) Turn off add-on visibility on your search heads.
3) Configure the search head tier to directly forward data to the indexer tier.
4) Distribute the summary index configurations to the indexer.
5) Install the add-on to a heavy forwarder.

I am little confused about step 3 and 5. Aren't they redundant?

If the search head can forward the data to indexers, why is there a need for a heavy forwarder?
Is the heavy forwarder used to take some 'load' off the indexer? In that case, shouldn't the search head forward the data read from AWS to the heavy forwarder rather than the indexer?

[deepashri_123]

Hey@sreerajms,

You can add both, the app and add-on on the search head and configure the search head to forward logs to indexer.
https://docs.splunk.com/Documentation/Splunk/7.0.3/DistSearch/Forwardsearchheaddata
There is no need for heavy forwarder.

Let me know if this helps!!

[sreerajms]

@deepashri_123. Thank you. But I am not looking to bypass the Heavy Forwarder. We do plan to set up the Heavy forwarder, but was wondering why the search head would need to look at the AWS directly when there is heavey forwarder that can look into AWS logs and forward it to the indexer.

[deepashri_123]

In that case you can configure add-on on heavy forwarder and forward data to indexer and aws-app on search head.

[sreerajms]

Did you mean do not install the add-on on the search head?

[deepashri_123]

yes!! You need add-on only on one server.

[sreerajms]

Thanks Deepa. I'll try that and let you know.

[klaxdal]

The heavy forwarder facilitates the connection / data parsing but does not 'index'

Additionally the data is not read from AWS by the search head but rather the search head is looking at the indexed data . There are some components which require the search head to forward data to the indexer .

[sreerajms]

Agreed. Any idea why we need to configure Search Head with AWS access details and configure forwarding to the Indexer when there is a heavy forwarder that can forward the data to the indexer?