Last checked and updated on April 15, 2026What Is New in Elasticsearch 7.14
Elasticsearch 7.14 delivers significant enhancements in data management, search performance, and operational resilience. This release focuses on making complex operations more efficient and user-friendly for developers and operators.
| Category | Key Updates |
|---|---|
| New Features | Searchable Snapshots GA, Failover for Cross-Cluster Replication, geohex_grid Aggregation |
| Enhancements | Faster Recovery, Improved Logging, Better CCS Performance, Data Streams Lifecycle Management |
| Security | FIPS 140-2 Compliance, Service Account Token Management |
| Deprecations | Deprecation of node-level search.remote settings |
How does Searchable Snapshots improve my storage strategy?
Searchable Snapshots have moved to General Availability, allowing you to mount snapshot indices directly as searchable entities. This feature fundamentally changes cold and warm storage tiers by letting you search data directly from your object store like S3.
In practice, this means you can reduce your local storage footprint significantly while keeping all your data searchable. The performance is slower than fully local indices, but it's a game-changer for archival data that still needs occasional querying.
What's new for Cross-Cluster Replication reliability?
Automatic follower index failover is now available for Cross-Cluster Replication (CCR). If the remote leader cluster becomes unavailable, the follower indices will automatically pause and attempt to resume once the connection is re-established.
This removes a major operational headache. Previously, a network partition could leave follower indices in a stuck state requiring manual intervention. Now the system handles this gracefully, which is crucial for maintaining replication across availability zones.
Are there new geo-analytics capabilities?
The new geohex_grid aggregation enables Uber's H3 hexagon-based geospatial aggregations directly in Elasticsearch. This provides a more natural way to analyze and visualize geospatial data compared to traditional rectangular grids.
Hexagons offer better visualization and more accurate spatial analysis since they reduce sampling bias at the edges. You can now create hexbin maps and perform spatial aggregations without needing to post-process your data externally.
How is cluster recovery performance improved?
File-based recovery operations are now significantly faster due to optimizations in how Elasticsearch handles file chunks. The system processes larger chunks more efficiently during peer recovery and snapshot restore operations.
This translates to reduced downtime when restoring from snapshots or when nodes join the cluster and need to sync data. For large clusters with big indices, this improvement can cut recovery times noticeably.
What security enhancements should I know about?
Elasticsearch 7.14 achieves FIPS 140-2 compliance when running on a FIPS-enabled JVM, meeting stricter cryptographic standards required for government and financial sectors. Additionally, service account tokens now support manual invalidation.
The FIPS compliance means you can deploy Elasticsearch in environments with strict cryptographic requirements without custom modifications. The token invalidation gives operators better control over security credentials for automated processes.
FAQ
Can I use Searchable Snapshots with my existing snapshots?
Yes, you can mount any existing snapshot repository that uses a supported object store (like S3, GCS, or Azure Blob Storage). The feature works with both new and previously created snapshots.
Does the CCR failover work for bidirectional replication?
The automatic failover handles unidirectional replication scenarios. For bidirectional setups, you'll need to configure failover on both clusters independently since each replication stream is managed separately.
What's the performance impact of using geohex_grid versus geotile_grid?
The geohex_grid aggregation has similar performance characteristics to geotile_grid but provides more accurate spatial representations. For most use cases, the difference is negligible unless you're aggregating over extremely large datasets.
Do I need to change my snapshot repository configuration for Searchable Snapshots?
No, your existing repository configuration remains compatible. The searchable snapshot functionality builds on the same repository infrastructure that already exists in Elasticsearch.
How does FIPS mode affect my cluster's performance?
FIPS-compliant cryptographic operations may have a slight performance overhead compared to standard implementations, but for most workloads the difference is minimal. The trade-off is acceptable for environments that require FIPS certification.