Optimum payload size for ingesting events via splunk HEC

[kkarthik_splunk]

I am trying to ingest data into Splunk via Splunk HEC using a python script. I am also sending the data in batches.

What should be the optimum size of the payload(data) that can be sent in a single post request to optimize the performance of the ingestion script?

[gjanders]

There is a Splunk conf presentation that covers this but I'm unsure which one. It will be on https://conf.splunk.com/watch/conf-online.html