SolrCloud never creates replicas automatically, unless perhaps you’re using the HDFS-only autoAddReplicas option. Start the new node using the same ZK, and then use the Collections API (https://cwiki.apache.org/confluence/display/solr/Collections+API) to ADDREPLICA.
The replicationFactor you specified at collection creation is pretty much ignored after creation. On 5/25/16, 8:53 AM, "Scott Chu" <scott....@udngroup.com> wrote: >I start 2 nodes and a zk ensemble to manager these nodes. Then I create a >collection with numShards=1 and replicationFactor=2 on 1st node. It spread >onto 2 nodes (meaning 1 leader, 1 replica). Now I want to add 3rd node but >don't do splitsharding. Before I try I want to ask some questions: Do I just >start node3 and join in same zk ensemble, Solrcloud will automatically create >replica on 3rd node? Or do I have to manually call some API to add replica to >3rd node? Either way, doesn't this exceed the replicationFactor? > >Scott Chu，scott....@udngroup.com >2016/5/25 (週三)
