The Gluster Community today announced the general availability of GlusterFS 3.7 open software-defined storage software. Feature-packed with performance and data protection capabilities, GlusterFS 3.7 delivers effective storage in data-intensive, high availability (HA) enterprise IT environments.
GlusterFS 3.7 includes new data protection features, such as:
- Bitrot detection identifies disk errors that silently corrupt data, by signing all files/objects in the volume, verifying data periodically, and logging anomalies.
- Erasure-coded volumes provide a flexible, durable alternative to traditional replicated volumes for capacity-optimized use cases.
GlusterFS 3.7 also includes several new performance and data availability features, such as:
- Advanced NFS support including Parallel NFS (pNFS) and Network File System version 4 (NFSv4) for Gluster volumes, including NFSv4 delegations (experimental feature) and lease locks, reducing client-server interactions, and enabling increased file operation performance in many scenarios.
- Highly-available NFS with NFSv4 delivers enhanced resource monitoring and active-active failover capabilities using NFS-Ganesha.
- Performance enhancements based on a new multi-threaded model results in overall reduced re-balance times at large scale and improved storage cluster elasticity. Multi-thread events handle metadata-intensive use cases, including small file workloads and various unstructured data use cases.
- Efficient object count underpins the inode quota feature. Optimized lookup of directory object count enables an efficient quota capability based on number of files in a directory hierarchy.
- Volume tiering (experimental feature) provides policy-based movement of data between fast and slow media tiers, enhancing performance for many small file workloads. Default configuration provides cache tiering for hot files automatically based on file access patterns.