On Wed, Jan 23, 2013 at 6:15 PM, Markus Jelsma <markus.jel...@openindex.io> wrote: > We need, and i think many SolrCloud users are going to need this as well, to > make replica's don't deviate too much from eachother, because if they do > documents are certainly going to jump positions.
The synchronization that would be needed to guarantee identical lucene indexes would be prohibitively expensive. Perhaps the best solution would be to route users to the same replicas. A solr request could request a token that when resubmitted with a follow-up request would result in hitting the same replicas if possible. -Yonik http://lucidworks.com
