Cloud-Native Active-Active NVMe-oF
MayaScale delivers enterprise-grade block storage on AWS, Azure, and Google Cloud with performance that rivals direct-attached NVMe.
Pool local NVMe storage using NVMe-over-Fabrics (NVMe-oF) protocol to create shared block volumes with sub-millisecond latency and 99.99% availability.
Why MayaScale?
139μs Latency
Sub-millisecond read and write latency on Azure L-series, AWS i4i, and GCP N2 instances. 10-50x faster than traditional cloud storage.
Active-Active HA
Both nodes serve I/O simultaneously. Sub-second failover with no downtime. Industry-first Active-Active MDS+OSS design for block storage.
90% Cost Savings
No expensive SAN infrastructure required. $10K starting price vs $200K+ traditional solutions. Pay only for cloud instances you use.
Choose Your Cloud
MayaScale delivers the same sub-millisecond performance and Active-Active HA across all major cloud providers.
MayaScale on Azure
Deploy on Azure Laosv4, Lasv4, and Lsv4 instances with local NVMe drives. 139μs read latency validated.
- Azure Marketplace
- ARM Templates
- SQL Server, Oracle HA
MayaScale on AWS
Pool local NVMe from i3, i3en, i4i, and i4g instances. 10x faster than EBS with synchronous replication.
- AWS Marketplace
- CloudFormation
- RDS Migration
MayaScale on GCP
Leverage local SSDs on N2, N2D, and Z3 instances. 10x faster than Persistent Disk with multi-zone HA.
- GCP Marketplace
- Deployment Manager
- Cloud SQL Migration
Perfect For These Workloads
Mission-Critical Databases
SQL Server, Oracle, PostgreSQL, MySQL with 99.99% uptime and sub-millisecond latency. Active-Active HA eliminates single points of failure.
Virtual Machines & VDI
VMware, Hyper-V, Azure Virtual Desktop with shared boot LUNs. Fast boot times and VM mobility with NVMe-oF storage.
NoSQL & Real-Time Analytics
Cassandra, MongoDB, ScyllaDB, Redis with ultra-low latency requirements. High IOPS for distributed database workloads.
AI/ML & HPC
Fast data loading for GPU training, high-frequency trading where microseconds matter, and real-time inference workloads.
Enterprise Features
NVMe-over-Fabrics
NVMe-oF over TCP for cloud deployments. Also supports RDMA (iWARP, RoCE, InfiniBand) for on-premises.
Active-Active HA
Both nodes serve I/O simultaneously. Sub-second failover with no downtime or data loss.
Data Services
Thin provisioning, compression, deduplication, and synchronous mirroring built-in.
Standard Protocols
NVMe-oF and iSCSI support. Works with standard Linux, Windows (StarWind), and VMware initiators.
Multi-Cloud
Same architecture and management across AWS, Azure, and GCP. True cloud portability.
Easy Management
CLI and web console for management. Infrastructure-as-Code deployment with Terraform.
On-Premises and Cloud Deployments
NVMe-oF On-Premises
Configure and deploy NVMe-over-Fabrics SAN infrastructure at rack scale in your data center by pooling heterogeneous NVMe devices over RDMA capable interconnect:
- RDMA Protocols: RoCE, iWARP, or InfiniBand
- Maximum Performance: <100μs latency with RDMA
- Consolidation: Pool diverse NVMe devices into shared storage
- Ease of Use: CLI and web console management
NVMe-oF On-Cloud
Experience NVMe-over-Fabrics on AWS, Azure, and GCP using standard TCP/IP protocol without specialized RDMA hardware:
- NVMe-over-TCP: Cloud-native without RDMA/SR-IOV
- Sub-Millisecond: 139μs latency validated on Azure
- Local NVMe Pooling: Transform ephemeral storage to persistent
- Marketplace: One-click deployment on all major clouds
Supported Clients
Linux Distributions
- CentOS 7.9+ (kernel 3.10.0.1160+)
- Rocky Linux 8/9
- Red Hat Enterprise Linux 8/9
- Ubuntu 20.04/22.04 LTS
- SUSE Linux Enterprise Server
Windows & VMware
- Windows Server 2016+ (StarWind NVMe-oF Initiator)
- VMware ESXi 6.7+ (NVMe-oF initiator)
- Hyper-V 2019+ (iSCSI or NVMe-oF)
Contact support for other distributions and platforms.
Ready to Deploy MayaScale?
Get started in minutes on AWS, Azure, or GCP. Deploy from marketplace or contact our team for pricing and custom configurations.
Try Now Contact Sales