WekaIO Matrix software is the industry’s first flash-native parallel file system that delivers unmatched performance to the most demanding applications, scaling to Exabytes of data in a single namespace.
Want to get the fastest performance to your AI and technical compute applications that need a shared storage platform? WekaIO Matrix is an easy to deploy software-based, scale-out storage solution that provides all flash performance on NVMe, SAS or SATA storage. Run our software on commodity server infrastructure; on-premises, in the public cloud, or hybrid and get NAS simplicity and manageability, cloud scalability, and breakthrough economics.Learn More
Predictable and seamless performance scaling is a challenge with traditional NAS filers and scale-out file systems, especially for workloads with small files or random I/O. Matrix gives you fine-grained control to dynamically scale performance independent of capacity based on application needs. WekaIO can easily deliver over 4GBytes/second to a single GPU or CPU based server and scale performance across the cluster. The software can be deployed as a dedicated storage appliance or hyperconverged on Ethernet and InfiniBand networks.
Matrix eliminates the infrastructure complexity, underutilized capacity and management nightmares created by legacy parallel file system architectures. The software-only solution is an NVME-native, fully parallel and distributed, POSIX compliant file system designed from the ground up to scale to thousands of compute nodes and petabytes of storage. Patented distributed data protection and integrated tiering to any S3 compatible object based cloud ensure your data is safe and always accessible. Maintenance and management is easy through the intuitive Trinity software console.
WekaIO's hardware independent architecture can be deployed on-premises or in the cloud on any commodity X86-based hardware. Run it hyperconverged with the applications or on dedicated storage servers.
Feature-for-feature, WekaIO offers more when compared to other file system solutions.
WekaIO is the first hardware independent, shared accelerated storage platform that turns any Intel-based commodity server into a high-performance, scale out storage system.
An NVMe-native, resilient and coherent, POSIX compliant file system with integrated tiering to object storage for best capacity scaling and cloud economics
An easy to use, intuitive management console that makes deploying, managing, and scaling your storage cloud radically simple
An integrated cloud-bursting solution for elastic GPU and CPU compute scaling and limitless storage capacity
|Architecture||Intel x86 standard server - 1 core minimum and 10GBytes RAM|
|Cluster Size||6 Servers (minimum failure domain) to 1,000's in a single cluster|
|Data Protection||Distributed data protection (Data+Parity of N+2 or N+4)|
|Supported Protocols||POSIX, NFS, SMB, HDFS|
|Maximum Number of File Systems||Architectural limit 64,000, tested limit 1,024|
|Storage Types||Enterprise class SSD – SATA, SAS, PCIe or NVMe|
|Deployment Method||VM (KVM, ESXi, XEN), Bare metal process, AWS instance|
|Management Interface||GUI, CLI, REST API|
|Snapshots and Clones||File and file system level, up to 1,024 snapshots|
|Snapshot to S3 Object Storage||Backup to the cloud with single software, burst to the cloud for on-demand compute resources|
|Deployment Models||Hyperconverged (on-premises or AWS) or dedicated storage server|
|Error Detection||End-to-end data protection|
|System Monitoring||Cloud based monitoring and analytics for application tuning|
|System Alerts||Configurable to Email, SMS, Slack|
|Access||Single sign-on integration|
|Networking||10GbE or higher|
|Operating Systems||Linux – RHEL 6.5 or later, CentOS 6.6,6.7,6.8, Ubuntu 14.0.4, SUSE 11.4|
|Cloud Tiering||S3, REST, public or private cloud, up to 8 unique targets|
|Certified Object Stores||Amazon, HGST, Cleversafe, Cloudian|
|IOPS Performance (NVMe)||4K Reads - 50,175 at <200 microsecond latency|
|Throughput Performance (NVMe)||1MByte Reads - 980MBytes/second|
Register for your free copy now.