In order to configure SolrCloud to run with large indexes, it is important to first design the system based on the requirements. The design has to be based on the following factors:
Number of nodes participating in the cloud
Distribution of shards and their replicas over nodes
Replication factors and leader
ZooKeeper setup
Prerequisites for this would require Apache Solr, ZooKeeper, J2EE container (optional).
First, we need to set up a ZooKeeper ensemble on all the nodes. Although Apache Solr ships with embedded ZooKeeper, for large indexes and scalability requirements, it is recommended to go ahead with a full ZooKeeper set up. You can download the latest version of Apache ZooKeeper. Now, unzip the download on all the nodes, and edit the zoo.cfg
file in your ZKHOME
| conf
folder; in that file, you need to specify the list of ZooKeeper servers as shown in the following screenshot. You must also specify correct dataDir...