Elasticsearch Interview Questions: Architecture, Indexing, Optimization, and Operations

1. Elasticsearch Overview and Cluster Architecture

Interview question about understanding ES usage, cluster size, index data volume, shard count, and tuning methods. Example answer mentions a 13‑node cluster, multiple indices, daily data growth, and shard configuration.

1.1 Design‑time Optimizations

Use date‑based index templates with rollover API, alias management, nightly force_merge, hot‑cold data separation, curator for lifecycle, appropriate analyzers, and thoughtful mapping.

1.2 Write‑time Optimizations

Set replica count to 0 before bulk writes, disable refresh_interval, use bulk API, restore replicas and refresh after write, and prefer auto‑generated IDs.

1.3 Query Optimizations

Disable wildcard and large terms queries, use keyword fields when possible, limit searches to time‑based indices, and configure routing.

1.4 Other Optimizations

Deployment and business‑level tuning.

2. Inverted Index

Explanation of inverted index as a term‑to‑document mapping, enabling O(1) lookup, built on Lucene’s FST structure for low space usage and fast queries.

3. Handling Large Indexes

Discusses dynamic index creation with templates and rollover, cold‑hot storage separation, and scaling the cluster by adding nodes without restart.

4. Master Election

Only nodes with master: true can be elected; min_master_nodes prevents split‑brain; election chooses node with smallest ID.

GET /_cat/nodes?v&h=ip,port,heapPercent,heapMax,id,name

5. Document Indexing Process

Describes client request routing, shard allocation, primary write, replication to replica shards, and acknowledgment flow.

shard = hash(_routing) % (num_of_primary_shards)

6. Search Process

Query‑then‑fetch: query phase runs on each shard, builds local priority queue, merges globally, then fetch phase retrieves full documents.

7. Linux Tuning for ES

Disable swap, set JVM heap to half of RAM (max 32 GB), increase max file handles, adjust thread‑pool sizes, and use RAID10 storage.

8. Lucene Internals

Brief overview of Lucene’s indexing and search architecture.

Conclusion

Emphasizes deep understanding and practice for interview success.