I've added a new indexer to the cluster and I would like to force replication of some older archived indexes. I'm curious if there's a way to trigger Splunk to replicate the indexes, since they are not being written to anymore, or is it better for me to just copy the index files over?
Also, any best practices on configuration for an archived index to ensure it's not wasting resources.
Thanks in advance,
David
If its an older archive that was not generated from an active cluster member, there's not much you can do to force it to replicate. Index buckets created pre-clustering aren't in the right format to be replicated. If you want to, you could distribute the buckets to other cluster members, but that won't give you redundancy, it will just distribute the search load. You have to make sure to avoid bucket numbering collisions as well as duplicating buckets across the cluster (or you'll see duplicate events in your search).
If its an older archive that was not generated from an active cluster member, there's not much you can do to force it to replicate. Index buckets created pre-clustering aren't in the right format to be replicated. If you want to, you could distribute the buckets to other cluster members, but that won't give you redundancy, it will just distribute the search load. You have to make sure to avoid bucket numbering collisions as well as duplicating buckets across the cluster (or you'll see duplicate events in your search).
That's what I thought, thanks so much for confirming.
Could you clarify what you mean by "archived index"?
Sorry, my definition is older data that was moved from old version of splunk, configured to be read from but no longer written to. the readonly option is set against the index.