I was reading here: http://wiki.apache.org/solr/NewSolrCloudDesign There says something about:
*Split_partition* : (params : partitionoptional). The partition is split into two halves. If the partition parameter is not supplied, the partition with the largest number of documents is identified as the candidate. What is that partition split? Is it something like that that resolved issue: https://issues.apache.org/jira/browse/SOLR-3755
