What is the difference between using repFactor=0 versus repFactor=1?

[goodsellt]

Can anyone explain to me the differences between using repFactor=0 and repFactor=1?

We only replicate specific critical indexes in our environment (in an interest of conserving disk space). However, by not having repFactor set (so it defaults to 0), those indexes do not appear in the list of indexes in the indexer clustering UI and are not eligible for things like data rebalancing. So going back to my original question, would we get the ability to monitor the status and rebalance the data for these non replicated indexes if we set repFactor to 1 or is this just intended behavior of Splunk currently?

[Steve_G_ · Accepted solution]

The repFactor setting has one function only: It determines whether an index gets replicated across the cluster. All clustered indexes must include the setting repFactor=auto, and this setting must be configured identically on all the peer nodes. As with other index-related settings, the way to do this is to distribute the updated indexes.conf file via the configuration bundle methd.

For details, see http://docs.splunk.com/Documentation/Splunk/6.5.1/Indexer/Configurethepeerindexes#The_indexes.conf_r...

Note: The indexes.conf.spec file implies that repFactor can be set to multiple values. At this time, the only valid values for repFactor are 0 and auto.

[goodsellt]

Thanks for the clarification on this! Yes I was looking at the spec file primarily and that's where I started getting a bit confused by the description of the attribute.