Deployment Architecture
Highlighted

In a multisite indexer cluster, how can we verify that each peer does not get more than one copy?

I have set up a Splunk indexer cluster on AWS using the multisite feature. The cluster is spread across 2 AZs. Each AZ has 2 peers.
My site rep factor is: origin:2, site1:2, site2:2, total:4.
I'm assuming that each peer will get only 1 copy of data. Is this assumption right?
If so, how can I verify that each peer does not get more than one copy?

0 Karma
Highlighted

Re: In a multisite indexer cluster, how can we verify that each peer does not get more than one copy?

Splunk Employee
Splunk Employee

Yes, in all indexer clusltering, single-site or multisite, each peer gets at most a single copy.

You might find it useful to review the discussion of how clusters handle buckets across the set of peer nodes: http://docs.splunk.com/Documentation/Splunk/latest/Indexer/Bucketsandclusters

This discussion focuses primarily on single-site clusters, but the principles are the same for multisite clusters.