Web / Apache Solr Interview questions
Why is SolrCloud preferred over standalone Solr for production?
Standalone Solr with master-slave replication works, but it has operational gaps that matter at production scale:
- No automatic failover - if the master goes down, writes stop until someone manually promotes a slave.
- No automatic sharding - a single core is limited by one machine's disk and memory.
- Manual config sync - schema and solrconfig changes have to be copied to every node by hand.
SolrCloud solves each of these using ZooKeeper: it distributes a collection across shards for horizontal scale, replicates each shard for high availability, elects a new leader automatically if one fails, and stores a single shared config set so every node in the collection uses the same schema and settings. For anything beyond a small, low-traffic index, SolrCloud's operational simplicity generally outweighs its added ZooKeeper dependency.
What happens to writes in standalone master-slave Solr if the master fails?
What does SolrCloud use to keep every node's schema and config in sync?
More Related questions...
