Shard Count Calculator
Enter your total data size, node count, shard size limit, and replication factor to get the recommended shard count for your database cluster.
Cluster Parameters
Raw dataset size before replication.
Data nodes in your cluster (exclude coordinating/master-only nodes).
Elasticsearch recommends 10–50 GB. Commonly 20–30 GB.
1 = no replicas, 2 = one replica copy, 3 = two replica copies.
Fill in the parameters on the left and click
Calculate Shards to see results.
Recommended Shard Count
-
primary shards
Shards per Node
-
Avg Shard Size
-
Total Shards (w/ Replicas)
-
Storage per Node
-
Configuration Breakdown
Summary
Enter your total data size, node count, shard size limit, and replication factor to get the recommended shard count for your database cluster.
How it works
- Enter the total size of your dataset in gigabytes.
- Specify how many nodes (servers) are in your cluster.
- Set the maximum shard size — Elasticsearch recommends 10–50 GB per shard.
- Enter your replication factor (1 = no replicas, 2 = one replica per shard).
- The calculator computes the minimum shards needed to stay under the size limit, then rounds up to a multiple of your node count for balanced distribution.
- Review the per-node shard count and estimated shard size to validate the result.
Use cases
- Sizing a new Elasticsearch index before ingestion begins.
- Re-sharding an existing cluster after data growth.
- Planning shard allocation for MongoDB sharded clusters.
- Estimating hardware requirements for a given dataset and shard policy.
- Validating whether an existing shard count is appropriate for current data volume.
- Comparing configurations with different replication factors.
- Designing multi-tenant databases with per-tenant shard budgets.
- Capacity planning for time-series databases with rolling retention.
Frequently Asked Questions
Last updated: 2026-06-11 ·
Reviewed by Nham Vu