STEELDOME TECHNOLOGIES

INTRODUCTION TO HYPERSERV

HyperSERV is a next-generation data platform that unifies StratiSTOR's scalable, high-performance storage with StratiSERV's enterprise-grade virtualization and orchestration.

EXECUTIVE SUMMARY

HyperSERV brings enterprise compute and distributed storage together on one StratiSYSTEM operating foundation. It combines StratiSERV virtualization, Kubernetes, and workload services with StratiSTOR block, file, object, and data-protection services on the same infrastructure.

Designed for organizations modernizing VMware, Nutanix, and traditional three-tier environments, HyperSERV offers the operational benefits of hyperconvergence without committing every future workload to one fixed hardware configuration. Supported industry-standard Intel and AMD platforms, networking, storage media, and accelerators can be selected to meet the application.

Virtual-machine high availability, migration, snapshots, backup, replication, resilient scale-out storage, and centralized management work within a consistent operational model. Kubernetes, private cloud, GPU-enabled computing, and supported AI inference services extend that foundation beyond conventional HCI.

HyperSERV can begin with compact edge infrastructure, scale into resilient enterprise clusters, or integrate dedicated StratiSTOR and StratiSERV resources when compute and data requirements grow differently. Converge where it makes sense. Expand without replacing the operating foundation.

SERVICE ARCHITECTURE

HyperSERV runs StratiSERV compute and StratiSTOR data services on the same supported nodes. Each node can contribute processing, memory, local storage, networking, and—where configured—GPU resources while participating in shared management, availability, protection, and lifecycle operations.

Virtual machines and containers use the storage and networking services appropriate to their workloads. The integrated data layer can provide supported iSCSI and NVMe-oF block access, SMB and NFS file services, and S3-compatible object storage, with additional services enabled when required.

Compute, storage protection, Kubernetes, networking, and workload placement retain their own operational controls even when deployed together. Capacity, performance, redundancy, and resource reservations are therefore configured for the applications rather than imposed by a fixed appliance profile.

HyperSERV can also integrate dedicated storage clusters, additional compute-only nodes, compatible external storage, and specialized GPU infrastructure within broader StratiSYSTEM designs. Hyperconvergence is a deployment choice, not an architectural boundary.

HyperSERV converged compute and distributed storage foundation
HyperSERV brings StratiSERV compute and Kubernetes together with StratiSTOR block, file, object, and protected storage services on a common StratiSYSTEM operating foundation.

HYPERCONVERGED INFRASTRUCTURE FOR MODERN WORKLOADS

Modern HCI must do more than place a hypervisor and storage on the same servers. Enterprise applications now coexist with containers, AI services, large datasets, private-cloud platforms, and distributed sites—all with different performance and recovery requirements.

HyperSERV addresses these needs by converging essential infrastructure services while retaining the flexibility to separate or specialize resources when workloads demand it.

VIRTUALIZATION MODERNIZATION

Replacing an established virtualization or HCI platform involves more than moving virtual-machine disks. Availability, backup, recovery, networking, hardware compatibility, and day-2 administration all need to move with the workloads.

HyperSERV brings enterprise VM lifecycle management, supported live and storage migration, HA restart, templates, snapshots, backup, and replication together with integrated storage. This creates a practical modernization path without tying compute and storage to a single proprietary appliance lifecycle.

RAPID DATA GROWTH

Virtual machines, databases, backups, application content, and AI datasets continue to increase capacity and performance demand. Fixed configurations can force premature appliance refreshes or leave compute and storage resources poorly balanced.

HyperSERV expands by adding qualified nodes and storage devices. Where requirements diverge, the wider StratiSYSTEM architecture can incorporate dedicated storage or compute resources rather than requiring every expansion to increase both equally.

CLOUD-NATIVE AND MIXED APPLICATION ENVIRONMENTS

Established business applications remain on virtual machines while newer services increasingly use Kubernetes and containers. Operating these as isolated infrastructure platforms increases complexity and fragments resources.

Kubernetes is included through StratiSERV, allowing VM and container infrastructure to share a common compute, networking, monitoring, and storage foundation. OpenShift and other application platforms can be deployed on supported configurations when their additional services are required.

AI AND ACCELERATED INFRASTRUCTURE

AI, analytics, rendering, and research combine demanding CPU or GPU workloads with large model files, datasets, checkpoints, and fast-moving storage traffic. Keeping these resources in compatible infrastructure can simplify data access and operations.

HyperSERV supports qualified NVIDIA and AMD GPU options, passthrough, compatible virtual GPU technologies, high-speed networking, and StratiSTOR data services. Optional self-hosted inference deployments can use CPU or supported GPU resources, with local model APIs, shared multi-model web access, and knowledge retrieval where configured.

HIGH-SPEED NETWORKING

Converged nodes may simultaneously carry virtual-machine traffic, Kubernetes services, storage replication, migration, backup, and recovery flows. Network design is therefore central to the capacity and failure model.

HyperSERV can use high-speed Ethernet, bonding, VLANs, virtual networking, and selected SR-IOV, RDMA/RoCE, DPDK, VPP, and compatible offload technologies. Supported fabrics and separate traffic domains are chosen according to throughput, latency, security, and redundancy requirements.

DISTRIBUTED AND MULTI-SITE INFRASTRUCTURE

Applications increasingly span primary and secondary data centers, remote offices, edge sites, and service-provider environments. Resilience may require more than recovery from an individual host or drive failure.

HyperSERV can participate in multi-site infrastructure using qualified StratiSTOR protection designs, workload backup or replication, and appropriate network and quorum architectures. Synchronous protection can provide zero RPO for acknowledged writes while full cross-site protection is intact; service recovery time depends on the deployed workload and failover design.

OPERATIONAL SIMPLICITY AND RESOURCE FLEXIBILITY

Combining compute and storage reduces the number of separately managed infrastructure stacks, but effective HCI still requires clear insight into CPU, memory, storage, network, and recovery headroom.

HyperSERV uses the shared StratiSYSTEM management and lifecycle model while allowing resource policy and deployment topology to evolve as workloads change.

INTEGRATED COMPUTE AND STORAGE SERVICES

HyperSERV delivers the capabilities of StratiSERV and StratiSTOR together. The services below are selected according to the deployment role, hardware, and protection model rather than enabled indiscriminately.

ENTERPRISE VIRTUALIZATION AND VM LIFECYCLE

Provision, configure, start, stop, monitor, protect, migrate, and retire virtual machines through StratiSERV management. Supported guest options include UEFI, Secure Boot, virtual TPM, virtual networking, device assignment, and compatibility settings for eligible host generations.

HIGH AVAILABILITY AND LIVE MIGRATION

Configured HA can restart protected VMs on eligible surviving hosts following supported failures, subject to shared storage access, remaining capacity, and placement policy. Live migration supports planned maintenance and workload movement between compatible hosts.

A converged design must reserve sufficient compute and storage capacity to remain operational during node failure and recovery. HA is not a substitute for application-level clustering or tested recovery procedures.

INTEGRATED BLOCK, FILE, AND OBJECT STORAGE

StratiSTOR provides distributed data services for VM storage, databases, file shares, backups, and cloud-native applications. Supported access methods include iSCSI, NVMe-oF, SMB, NFS, and S3-compatible object storage, with protection and service placement tailored to each workload.

DISTRIBUTED STORAGE PROTECTION

Replication and erasure coding can be selected to balance capacity efficiency, redundancy, write overhead, and failure tolerance. Policies can account for qualified device, node, rack, or site failure domains.

Storage health, recovery state, and free-capacity reserves must be assessed together; the presence of multiple nodes does not by itself guarantee a particular fault-tolerance level.

SNAPSHOTS, BACKUP, AND GRANULAR RECOVERY

HyperSERV brings VM snapshots and image-level backup together with StratiSTOR storage protection. Supported backup workflows can protect multiple virtual disks and recover complete VMs or selected guest files.

Granular restore workflows can address supported Linux and Windows filesystems, including encrypted guest volumes when the required recovery material and tooling are available. Backups remain a separate recovery control from snapshots or storage redundancy.

SCHEDULED VM REPLICATION AND DISASTER RECOVERY

VM replication enables protected workload copies at another approved destination for recovery, migration, and continuity planning. Scheduling, on-demand execution, destination handling, and eligible-VM controls help operators manage these workflows deliberately.

Recovery objectives depend on replication frequency, destination capacity, network bandwidth, and application consistency. Administrators should test restoration and recovery at the intended destination.

STORAGE INTEGRITY AND SELF-HEALING

Distributed checksums, background scrubbing, and recovery mechanisms help detect storage inconsistencies and restore data protection where valid redundant copies or fragments are available. Health monitoring also surfaces device issues, repairs, and placement-group recovery activity.

Not every integrity warning corrects itself immediately; persistent inconsistencies, repeated read repairs, or device errors require investigation and may call for operator intervention.

SDCACHE AND STORAGE ACCELERATION

Supported SDCache configurations use memory or disk-based acceleration to improve selected storage access patterns and provide a consistent device presentation layer. Cache allocation, write policy, and safeguards must be sized to the workload and the underlying media.

Write-back configurations require documented power-loss protection and recovery procedures; cache is not a substitute for resilient storage placement.

CAPACITY EFFICIENCY AND FLEXIBLE MEDIA

All-NVMe, SSD, HDD, hybrid, and supported external block designs can be combined with workload-specific replication, erasure coding, compression, and caching. This lets capacity and throughput follow application requirements instead of a single appliance bill of materials.

KUBERNETES, PRIVATE CLOUD, AND MODERN APPLICATIONS

HyperSERV supports traditional VMs and cloud-native workloads within a converged infrastructure model, reducing the need to create separate compute and storage islands for each application generation.

KUBERNETES AND CONTAINER INFRASTRUCTURE

Kubernetes services are available through StratiSERV. HyperSERV can provide the underlying compute and appropriate StratiSTOR-backed persistent storage for supported stateful applications, with network, storage-class, and application policies selected for the cluster.

OPENSTACK AND PRIVATE CLOUD

Supported OpenStack deployments can use StratiSERV compute and StratiSTOR persistent data services to build API-driven private-cloud environments. Tenant isolation, image services, virtual networking, and storage integration are configured according to the selected private-cloud architecture.

DEVELOPMENT, TEST, AND APPLICATION MODERNIZATION

VM templates, snapshots, Kubernetes, flexible storage, and infrastructure automation make HyperSERV suitable for development environments, CI/CD, isolated tests, and application modernization alongside production workloads.

AI, GPU, AND HIGH-PERFORMANCE COMPUTING

HyperSERV can bring accelerated compute and fast data access together, supporting AI and HPC use cases without requiring a completely separate operating model for the underlying infrastructure.

GPU PASSTHROUGH AND VIRTUAL GPU

Assign supported physical GPUs directly to VMs for dedicated acceleration or use compatible vendor vGPU technologies to share eligible devices where hardware, drivers, and licensing permit. GPU selection and placement must account for PCIe locality, memory, thermal limits, and workload mobility constraints.

SELF-HOSTED AI INFERENCE

Optional StratiSYSTEM AI deployment services can host compatible GGUF language models using CPU or supported GPU inference. Multiple local models can be exposed through a shared Open WebUI, with optional document retrieval, memory, search, and authenticated application access.

This complements VM and Kubernetes workloads; model sizes, concurrency, storage throughput, and compute resources must be sized explicitly rather than inferred from the presence of a GPU.

HIGH-PERFORMANCE DATA PIPELINES

StratiSTOR storage, qualified high-speed networking, supported GPU-aware data paths, and optional Lustre services can provide the data foundation for model files, training datasets, checkpoints, simulation, and large scientific workloads.

Actual GPUDirect and storage-path capabilities depend on the certified hardware/software combination and the data service in use.

HIGH-SPEED NETWORK AND ACCELERATOR INTEGRATION

Supported environments can use 25/100/200/400GbE and other qualified fabrics, alongside SR-IOV, RDMA/RoCE, selected DPDK/VPP data paths, and compatible DPU technologies. Workload performance depends on the complete CPU, PCIe, storage, NIC, and switch topology.

MULTI-SITE AVAILABILITY AND DATA PROTECTION

HyperSERV combines workload recovery and distributed storage options so customers can design for local faults, site events, or staged disaster recovery. Each level of protection must be qualified against its own failure and recovery assumptions.

SYNCHRONOUS DATA PROTECTION AND RECOVERY OBJECTIVES

Qualified StratiSTOR multi-site designs can acknowledge writes only after the required cross-site durability policy is satisfied. This can provide zero RPO for those acknowledged writes during full synchronous protection.

When a site is unavailable and a supported deployment continues in degraded operation, newly acknowledged writes may not retain the same cross-site protection until resynchronization completes. RPO and application recovery time are different objectives.

QUORUM, FAILURE DOMAINS, AND SPLIT-BRAIN PROTECTION

Site-aware designs use appropriate quorum and data placement to determine the authoritative surviving cluster membership. Two data-bearing sites may require an independent tiebreaker location to avoid competing writable partitions.

WORKLOAD-LEVEL RECOVERY AND REPLICATION

VM backup, scheduled replication, and tested recovery procedures complement storage-level redundancy. Administrators should document start order, dependencies, network changes, and workload validation following a site or cluster recovery.

SECURITY, ISOLATION, AND INFRASTRUCTURE CONTROL

HyperSERV uses supported StratiSYSTEM management, guest, and network controls to separate administration, data access, and application workloads across the converged platform.

SECURE BOOT, VIRTUAL TPM, AND GUEST PROTECTION

Supported UEFI Secure Boot and virtual TPM configurations provide security foundations for modern Windows and Linux virtual machines. Backup and recovery planning must account for virtual disks, VM configuration, boot firmware state, and TPM-associated material where required.

ROLE-BASED ADMINISTRATION AND TENANT ISOLATION

Role-aware access, network segmentation, and service-specific controls help separate infrastructure administrators, workload owners, and supported tenant environments. Access should be scoped to the actual services and data each role needs.

PRIVATE, SOVEREIGN, AND DISCONNECTED OPERATIONS

HyperSERV can support tightly controlled and disconnected infrastructure when the required local repositories, management services, time sources, certificates, and operational procedures are established. Sensitive workloads and data remain governed by the customer's chosen placement and access policies.

AUTOMATION, OBSERVABILITY, AND DAY-2 OPERATIONS

A converged platform must provide visibility across both compute and storage. HyperSERV combines the StratiSYSTEM operational view with management workflows for availability, data protection, upgrades, capacity, and expansion.

UNIFIED WORKLOAD AND STORAGE OBSERVABILITY

Administrators can review VM state, CPU and memory utilization, storage capacity, device health, recovery activity, network interfaces, and relevant service conditions from consistent management workflows. Correlating compute and data events shortens the path from a workload symptom to an infrastructure cause.

CAPACITY PLANNING AND RESOURCE HEADROOM

Hyperconverged sizing must account for VM and container resources, storage and cache memory, data protection overhead, rebuild bandwidth, and the resources needed to survive planned maintenance or a node failure.

Capacity planning should consider both physical storage consumption and workload demand; nominal raw disk capacity or vCPU totals alone are not a sufficient sizing basis.

POLICY, APIS, AND REPEATABLE OPERATIONS

Supported web, CLI, and API workflows allow provisioning, monitoring, VM lifecycle actions, protection tasks, and cluster maintenance to be integrated with broader automation and service-management processes.

MAINTENANCE AND ROLLING EXPANSION

Qualified node additions, storage expansion, migrations, and supported rolling maintenance can be performed while services remain available when redundancy and remaining resources permit. Health and protection state should be checked before and after each change.

COMMON HYPERSERV WORKLOADS

HyperSERV is intended for mixed environments where applications need compute, storage, networking, and protection under one consistent infrastructure operating model.

ENTERPRISE VIRTUALIZATION AND MODERNIZATION

Consolidate enterprise VMs, application servers, databases, and infrastructure services while gaining modern HA, migration, workload protection, and flexible hardware choices.

PRIVATE CLOUD AND SERVICE PROVIDERS

Build supported multi-tenant and API-driven private-cloud environments with combined compute, storage, network, and lifecycle services, whether for internal infrastructure or managed customer deployments.

VIRTUAL DESKTOP INFRASTRUCTURE

Host centralized virtual desktops with appropriate CPU, memory, storage, network, and compatible GPU resources while applying workload isolation and protection controls.

KUBERNETES AND CLOUD-NATIVE APPLICATIONS

Run containers and stateful Kubernetes applications on the same hardware used for enterprise virtualization, using approved networking and persistent-storage configurations.

AI INFERENCE, ANALYTICS, AND HPC

Combine capable CPU/GPU compute, local or distributed model and dataset storage, high-speed networking, and compatible specialized software for private AI, research, analytics, and simulation.

EDGE, BRANCH, AND REMOTE SITES

Use appropriately sized compact nodes or clusters for remote locations, laboratories, industrial sites, and branches while maintaining the same fundamental operations and management model used in the primary data center.

BACKUP, RECOVERY, AND MULTI-SITE SERVICES

Combine protected data services, VM backup and replication, and appropriately engineered site-level resilience for workloads that require defined recovery points and tested recovery procedures.

DEPLOYMENT MODELS

HyperSERV supports multiple physical designs so customers can match convergence, resiliency, cost, and growth to their workload instead of adopting one immutable node configuration.

SINGLE-NODE AND COMPACT INFRASTRUCTURE

Use small installations for labs, development, edge, and explicitly non-HA services. A single node does not provide node-level fault tolerance, regardless of the services installed.

HIGHLY AVAILABLE MULTI-NODE CLUSTERS

Use an appropriately sized multi-node architecture with validated quorum, storage protection, capacity headroom, network redundancy, and workload placement to support production availability objectives.

CONVERGED AND COMPUTE-ONLY EXPANSION

Add HyperSERV nodes when compute and storage demand grow together, or integrate supported compute-only StratiSERV resources when additional CPU, memory, or GPU capacity is needed without proportional storage growth.

DEDICATED STORAGE AND HYBRID ARCHITECTURES

Combine HyperSERV with dedicated StratiSTOR storage, separate StratiSERV compute, supported third-party storage, or specialized high-throughput systems where capacity and performance requirements diverge.

MULTI-SITE AND STRETCH ARCHITECTURES

Supported cross-site deployments can use qualified data protection, quorum, networking, and VM recovery designs to address geographic availability and disaster-recovery requirements.

DESIGNED FOR CONTINUOUS INFRASTRUCTURE EVOLUTION

The useful life of enterprise applications and data often exceeds the life of the physical servers they run on. HyperSERV separates the operating foundation from a fixed appliance refresh cycle.

Supported new server generations, storage media, network capabilities, and accelerators can be introduced as requirements evolve. Convergence remains an option, while dedicated compute or storage can be introduced when it better suits the workload.

Modernize the applications. Expand the infrastructure. Keep the operating model.

CONCLUSION

HyperSERV unifies StratiSERV compute and StratiSTOR distributed storage within a common software-defined environment for enterprise virtualization, cloud-native applications, private cloud, AI, and data-intensive workloads.

Its advantage is broader than traditional HCI: organizations can start with converged infrastructure, protect and move workloads, adopt new application models, and integrate specialized or disaggregated resources as requirements change.

One infrastructure operating foundation. Converged where it helps, flexible wherever the workload demands.