About real time search

[yutaka1005]

I want to know about CPU occupation when doing a real-time search.

If I build Splunk in a standalone way, and I configure a real-time search, I think that one of cpu core will be occupied.

But which server's cpu core is occupied by real-time search when configuring distributed search like indexer clustering?
will only cpu core of the search head be occupied? Or, because it is a distributed search, will cpu core of each search peer also be occupied?

Also, if I configured search head clustering, will cpu core of all members be occupied?

I am planning to create large scale configuration for personal use, and planning configure alerts using real time search (rolling window) in the environment, so I want to know how to use cpu core.

I appreciate if someone tell me about it.

[richgalloway · Accepted solution]

In a distributed real-time search, one core for each peer is occupied, but only one core on one search head is used.

[yutaka1005]

Thank you for answer.

you mean that only one core on one search head is used if search is processed in search head clustering?

[richgalloway]

Yes, that is what I meant.

[yutaka1005]

Thank you for answer!

I understood it!