Elasticsearch Scalability and Resilience:

• Elasticsearch is distributed by nature, ensuring high availability and scalability.
• Add servers (nodes) to a cluster to increase capacity; Elasticsearch distributes data and queries across nodes.
• No application overhaul needed; Elasticsearch balances multi-node clusters automatically.
• Under the hood, an index consists of one or more physical shards, each being a self-contained index.
• Shards distribute documents across nodes for redundancy and query capacity.
• Two types of shards: primaries (store documents) and replicas (copies of primaries).
• Replicas serve read requests (searching or retrieving documents).

Shard Considerations:

• Shard size and number impact performance.
• Aim for average shard size between a few GB and a few tens of GB.
• Avoid excessive shards per node (proportional to heap space).
• Test with your own data to find optimal configurations.

Cross-Cluster Replication (CCR):

• Synchronize indices from primary to secondary remote clusters.
• Secondary clusters serve as hot backups or read-only followers.
• Active-passive replication; primary handles write requests, secondary is read-only.
• Use CCR for disaster recovery and geo-proximity read requests.