Enterprise-Grade Private Cloud. Zero Vendor Lock-in.
We design, deploy, and operate OpenStack environments built for the long haul. Debian-native. Fully automated. Designed for long-term cost predictability and operational independence.
OpenStack, The Way It Should Be
Not every OpenStack deployment is the same. Ours is opinionated—because production environments demand clarity, not flexibility theater.
Debian-Native
We build on Debian stable. We avoid unnecessary abstractions that obscure operational visibility. Our focus is on Linux-native systems that engineers can troubleshoot confidently—even at 2 AM.
VM-First Architecture
Virtual machines remain the right abstraction for most enterprise workloads. We design OpenStack as a VM platform first—Kubernetes runs on top when you need it, not the other way around.
Built for Operations
Anyone can demo OpenStack. We build OpenStack that survives its first upgrade, its first security patch, its first capacity expansion. Architecture decisions on day one are driven by what happens on day 500.
Integrated by Design
Compute, storage, and networking are not separate conversations. Every deployment integrates Nova, Ceph, and OVN as a unified platform—not three vendors bolted together.
The Full Lifecycle, Not Just Day One
Most vendors disappear after installation. We stay because that is when the real work begins.
Adoption
Starting Right Means Operating Right
Migration
Exit VMware Without Disrupting Operations
Day-2 Operations
Where Most OpenStack Projects Fail. Where We Excel.
Customization
Your Platform, Your Brand, Your Workflows
For ISPs & Data Centers
Build Your Own Regional Cloud
Indian ISPs and data center operators increasingly want to offer “Local Cloud” services to their customers. We help you build that capability—with your brand, your pricing, your customer relationships.
Complete Cloud-in-a-Box
Full OpenStack deployment with Ceph storage, OVN networking, and customer-facing portal. Ready for multi-tenant operations.
White-Label Everything
Dashboard, API endpoints, documentation—all branded as your service. Your customers never interact with Padmini.
Billing Integration
Usage metering tied to your existing billing cycles. Support for prepaid, postpaid, and hybrid billing models.
Tiered Resource Allocation
Define your own VM flavors, storage tiers, and network packages. Price them however makes sense for your market.
Operational Handover
We train your NOC team to handle L1/L2 operations. We remain available for L3 escalations and platform upgrades.
Integrated Platform, Not Assembled Parts
OpenStack is the orchestration layer. Underneath, we integrate best-in-class open source components into a unified platform.
Padmini Cloud Stack (PCS)
Enterprise Open Source Infrastructure Platform
Security & Compliance Layer
Firewalls • Workload Isolation • Encryption • Access Control • Audit & Compliance
PCS Cloud Control Plane
Compute Management
Network Virtualization
Storage Orchestration
Identity & Access Management
Image & Lifecycle Management
API & Automation Services
Distributed Storage Services
Block • Object • File • Backup & Replication
Software-Defined Networking
Virtual Networks • BGP Routing • Segmentation • Firewall & Policy
Container Orchestration Platform
Kubernetes-based Workloads • Platform Services • Application Lifecycle
Debian-based Operating System Layer
Hardened OS • Custom Kernel • Performance Tuning
Bare Metal & Virtual Infrastructure
Physical Servers (x86 / ARM) • Virtual Machines • Storage Nodes • Network Devices
Migration & Platform Adoption
Legacy → PCS • Phased Migration • Zero / Minimal Downtime
Padmini Cloud Stack (PCS) is an enterprise-grade private cloud platform engineered by Padmini Systems using Debian-based open-source components. The architecture is illustrative and represents our customized implementation for compute, storage, networking, security, and operations.
Compute
Bare-metal hypervisor performance with enterprise VM management. Live migration, automated placement, resource overcommit controls. Tuned for your workload profile—whether dense VM packing or performance-isolated workloads.
Storage
Software-defined storage that scales horizontally. Block storage for VMs, S3-compatible object storage for applications, CephFS for shared filesystems. Single storage platform, multiple access methods, no proprietary appliances.
Networking
Software-defined networking with hardware-grade performance. Tenant isolation, security groups, floating IPs, load balancers. BGP integration for external connectivity—ISP peering, data center interconnect, hybrid cloud networking.
Security
Defense in depth across every layer. Network microsegmentation, API authentication, encrypted storage, audit logging. Designed for compliance with Indian regulatory frameworks—RBI, SEBI, IRDAI, MeitY guidelines.
Automation
Infrastructure as code from day one. Deployment automation, operational runbooks, self-healing workflows. Every manual process is a future automation candidate. We deliver infrastructure that gets easier to operate over time, not harder.
Proven Configurations
These are battle-tested configurations we have deployed. Use them as starting points for your planning.
Medium Enterprise Private Cloud
100-500 VMs, departmental cloud, development environments
Hardware Specification
- Control Plane3× servers (32 cores, 128GB RAM, 2× 480GB NVMe)
- Compute5-10× servers (64 cores, 512GB RAM, 2× 480GB NVMe)
- Storage5× servers (16 cores, 64GB RAM, 12× 4TB NVMe)
- Networking25GbE spine-leaf, 2× ToR switches per rack
Capabilities
- ✓~2,000 vCPU capacity (4:1 overcommit)
- ✓~200TB usable Ceph storage (3× replication)
- ✓Full HA control plane
- ✓OVN networking with BGP uplinks
Large Enterprise / Service Provider
500-2000+ VMs, multi-tenant, production workloads
Hardware Specification
- Control Plane3× servers (64 cores, 256GB RAM, 2× 960GB NVMe)
- Compute20-50× servers (128 cores, 1TB RAM, 2× 960GB NVMe)
- Storage10× servers (32 cores, 128GB RAM, 24× 8TB NVMe)
- Networking100GbE spine-leaf, dedicated management network
Capabilities
- ✓~10,000+ vCPU capacity
- ✓~1PB usable Ceph storage
- ✓Multi-site replication ready
- ✓Advanced networking (VPC, LBaaS, FWaaS)
Minimum Viable Production
Proof of concept, small workloads, learning environments
Hardware Specification
- Converged5× servers (32 cores, 128GB RAM, 2× 480GB + 4× 2TB NVMe)
- Networking10GbE, single ToR switch
Capabilities
- ✓~300 vCPU capacity
- ✓~20TB usable storage
- ✓HA control plane (3 of 5 nodes)
- ✓Basic networking
Note: This is the minimum for production. Below this, simpler virtualization solutions are more appropriate.
What Sets Us Apart
We are not a reseller. We are not a staffing company. We are infrastructure engineers who build and operate production platforms.
We Do Not Sell Licenses
No per-socket fees. No per-VM charges. No annual true-ups. You pay for our expertise in building and operating your platform. The software is open source—you own it completely.
We Operate What We Build
We do not hand over documentation and disappear. We stay engaged through Day-2 and beyond. Our incentive is long-term platform health, not project closure.
We Design for Failure
Every component fails eventually. We design assuming failure—redundant control planes, distributed storage, automated failover. Resilience is architecture, not hope.
We Automate Everything
If a task must be done twice, it becomes a script. If a script runs regularly, it becomes a service. We deliver platforms that become easier to operate over time.
We Understand Indian Enterprise
Compliance requirements, data sovereignty concerns, vendor negotiation dynamics—we operate in the same context you do. Local support, local understanding, aligned timezones.
We Contribute Upstream
We are not just consumers of open source. We contribute fixes, documentation, and features back to OpenStack and related projects. We understand the platform at source-code depth.
Our Philosophy
How we think about infrastructure shapes everything we build.
🔓Open Source First
We believe the best infrastructure is built on open source foundations. Not because it is cheaper—though it often is—but because it is inspectable, modifiable, and yours. No black boxes. No vendor lock-in. No licensing surprises.
⚡Automation as Standard
Manual operations do not scale. They introduce errors, create key-person dependencies, and make every change risky. We treat automation as the default delivery method—not a premium add-on.
😴Boring and Reliable
We choose technologies that are proven, stable, and well-understood. Innovation happens at the application layer. Infrastructure should be boring—predictably reliable, consistently available, never the reason for outages.
🏛️Sovereignty Matters
Your data, your hardware, your control. We help Indian enterprises build infrastructure that keeps sensitive workloads within their control—on-premises, in Indian data centers, under Indian jurisdiction.
📚Documentation is Delivery
If it is not documented, it is not delivered. Every platform we build includes operational documentation, architecture decisions, and runbooks. Your team can operate, troubleshoot, and extend without depending on our availability.
When OpenStack Is Not the Right Choice
We are transparent when OpenStack is not the right solution.
- ⚠Your team lacks dedicated infrastructure operations capacity and cannot commit to building it
- ⚠Your workloads are better suited to managed public cloud services (e.g., heavy use of managed databases, serverless, ML services)
- ⚠You need only a small number of VMs where simpler virtualization platforms would be more cost-effective
- ⚠You require capabilities OpenStack does not provide well (e.g., serverless-first architectures, edge deployments at hundreds of sites)
- ⚠You do not have (or cannot procure) server-grade hardware that meets minimum Ceph and OpenStack requirements—desktop or entry-level servers will not deliver production reliability
Our goal is long-term platform success, not forcing a technology fit. If OpenStack is not right for you, we will tell you—and can recommend alternatives that better match your situation.
Ready to Talk?
Whether you are planning a new private cloud, migrating from VMware, or struggling with an existing OpenStack deployment—we can help.
Request an Assessment
We will review your requirements, current environment, and goals. No obligation, no sales pressure—just a technical conversation about what is possible.
Get Started →Talk to Engineering
Direct conversation with engineers who will actually work on your platform. Not sales, not account managers—engineers.
Schedule a CallPlan Your VMware Exit
Dedicated resources for enterprises planning VMware migration. Assessment frameworks, cost comparisons, and transition guides.
Learn More →