LifeInCloud, a European public cloud provider, is on a mission to make enterprise-grade cloud solutions more accessible, without the complexity that often comes with them. Their philosophy is simple: powerful cloud infrastructure without the traditional headaches. As LifeInCloud embarked on building a scalable, high-performance public cloud, they turned to ShapeBlue as a leading expert in Apache CloudStack. With our deep expertise in CloudStack architecture, high availability, and automation, we helped LifeInCloud optimise its cloud platform—ensuring efficiency, reliability, and seamless scalability.

The cloud provider reached out to ShapeBlue during their transition. Their multi-datacenter requirements and focus on high availability demanded careful architectural planning and optimization.

Bogdan, LifeinCloud’s CEO, explained their decision: “Building a cloud platform isn’t a solo venture. ShapeBlue helped us nail the tricky bits, validated our architecture decisions and helped us optimize for future growth.”

George, LifeinCloud’s CTO, described the collaboration with ShapeBlue: “We needed proven solutions. When we got stuck with the HA configuration, their team had seen it before and knew exactly what to check.”

This case study explores their journey, the challenges they overcame, and how our collaboration helped them build a next-generation public cloud.

 

About LifeinCloud

LifeinCloud is a privately-owned, European public cloud provider on a mission to make powerful cloud solutions more accessible to developers and businesses worldwide. Or, as their website puts it, “enterprise-grade infrastructure, without enterprise-grade headaches” – their words, not ours. But let’s be honest, who wouldn’t want that?

This case study explores their journey of building a scalable cloud platform using Apache CloudStack (ACS) and how they’re delivering on that no-headache promise.

 

LifeinCloud’s Journey to Building a Public Cloud 

Like many companies in the industry, LifeinCloud started in 2009 as a managed services provider and later expanded into private cloud solutions.

But as demand grew, managing isolated environments for each client became a bottleneck: scaling was complex, infrastructure was underutilized, and onboarding took weeks.

 

“Each client environment was essentially a separate island. We were duplicating efforts across different environments – separate hardware, separate maintenance windows, separate upgrade cycles. This wasn’t just inefficient; it was unsustainable as we looked to scale.”

Bogdan Rohan, LifeinCloud Founder & CEO

 

 

To solve this, they built a multi-region public cloud platform from the ground up, powered by CloudStack – which transformed their operations:

• 40% reduction in overhead
• 85%+ infrastructure utilization
• onboarding time cut from weeks to hours
• eliminated full hardware refresh cycles in favor of incremental scaling

This move was also validated by the growing demand for open-source, vendor-neutral cloud solutions.

 

Why LifeinCloud Chose Open-Source

We all know about the acquisition that sent shockwaves through the cloud industry, forcing businesses to rethink their reliance on proprietary infrastructure.

LifeinCloud’s CEO, Bogdan, offers his insights:

“When you build on proprietary platforms, you’re ultimately at the mercy of whoever owns that platform. It doesn’t matter how stable the company seems today or how long you’ve been their customer. One board decision, one acquisition, one strategy shift can force you to either accept unfavorable terms, undertake a costly migration, or in some cases, go out of business.”

This wake-up call validated LifeinCloud’s decision to build their cloud on open-source technology, as Bogdan adds:

“We wanted to go a different route because we don’t like vendor lock-in. Our customers stay with us because they want to, not because they have to. However, if they decide to leave, we’re sad to see them go – but they can easily migrate to another open-source platform.”, he adds.

With their strategic direction clear, they turned to the technical evaluation. They needed a platform that could deliver on their vision of independence and simplicity while meeting enterprise requirements for reliability and performance.

 

Technical Deep Dive: How They Chose CloudStack 

Their technical evaluation came down to two major open-source contenders: OpenStack and Apache CloudStack. 

Let’s hear from LifeinCloud’s CTO, George Lisandru, who led the technical evaluation:

 

“OpenStack’s architecture follows the Unix philosophy – many small components each doing one thing well. In theory, this provides maximum flexibility. In practice, it means you need separate services for compute (Nova), networking (Neutron), identity (Keystone), image management (Glance) and storage (Cinder). Each component needs its own high-availability setup, monitoring and maintenance. It’s as powerful as it is complex.”

 

According to the provider, this complexity isn’t just an inconvenience – it has real operational implications:

1. Resource Overhead

OpenStack’s distributed architecture requires significant resources just to run the control plane. Each service needs its own compute resources, and the overhead grows as you scale. CloudStack’s centralized management approach is more efficient, requiring only a single management server (or a small HA cluster) regardless of infrastructure size.

2. Operational Complexity

Here’s a real-world example: setting up simple network functionality in OpenStack often requires configuring multiple Neutron plugins, ensuring they’re compatible, and maintaining them separately. With CloudStack, the same functionality is available out of the box through its integrated networking layer.

3. Storage Integration

LifeinCloud’s CTO explains a critical difference: “We tested both platforms with Ceph storage. OpenStack required additional components and careful configuration to work properly with Ceph. With CloudStack, the integration was native – it just worked. This pattern repeated itself across multiple aspects of the platforms.”

Feature	OpenStack	CloudStack
Architecture	Modular, requiring multiple services (Nova, Neutron, Keystone etc.)	Centralized management with built-in orchestration
Ease of Use	Requires a dedicated team for configuration and maintenance	Simpler to deploy and manage
Networking	Requires extensive Neutron configurations and plugins	Native SDN and VLAN support out of the box
Storage Integration	Needs additional services for Ceph compatibility	Native Ceph integration
Resource Requirements	High – demands multiple control nodes	Lower – single management server can handle orchestration

 

Resource Requirements at Scale 

The differences between platforms became clear when they tested real-world deployments. For this test, they ran 100 VMs to see just how much overhead each platform needed:

OpenStack needed 5+ control nodes and nearly 100GB of RAM just to manage itself. Each component – compute, storage, networking – needed its own high-availability setup. More components meant more points of failure.

CloudStack ran the same workload with a single management server (two for HA) and just 16GB of RAM. One control plane, one place to monitor, one system to maintain.

Performance Metrics

Their performance testing revealed significant improvements, as reported by LifeinCloud:

Compute Performance:

– VM Provisioning: 45 seconds average (down from 3+ minutes)

– Live Migration: Sub-second downtime for running VMs

– CPU Overhead: Less than 2% for KVM virtualization

Storage Performance:

– Random 4K Read: 120,000 IOPS per storage node

– Sequential Read: 3.2 GB/s sustained throughput

– Latency: Sub-5ms for cached reads

Storage Architecture: Why Ceph Won

The choice of Ceph as their storage platform wasn’t automatic. Their CTO, George Lisandru, provides insight into the decision: “We considered Linbit DRBD because it’s resource-efficient. But Ceph – while more resource-intensive – met our reliability requirements and offered better scaling characteristics.”

LifeinCloud’s current Ceph implementation provides:

– Triple replication for data redundancy

– Dedicated networks for replication traffic

– Optimized placement groups for better performance

– Integration with NVMe SSD storage for high-performance workloads

Networking: The Often-Overlooked Critical Component 

George highlights redundancy and isolation as their core networking principles. LifeinCloud built their network architecture with:

1. Full Network Redundancy

– Dual physical switches for hardware redundancy

– Network bonding on all hosts

– Automatic failover capabilities

2. Tenant Isolation

– VLAN-based network separation

– Granular firewall controls

– Support for custom network topologies

Security: Built-in, Not Bolted-on 

LifeinCloud wanted to build security as a fundamental component into every layer of their stack.

At the network level, CloudStack’s VLAN capabilities keep customer traffic completely separate – no customer can peek into another’s network space. Each VLAN gets its own firewall rules, so customers can lock down their environment exactly how they want it.

When it comes to access control, they implemented a granular Role-Based Access Control system where users get exactly the permissions they need – nothing more. Every admin action gets logged, not just for security, but as a critical requirement for their European customers who need to prove compliance with data protection laws.

Day-to-Day Operations

Running cloud infrastructure at scale reveals truths that weren’t apparent during LifeinCloud’s initial platform evaluation. The centralized management approach proved invaluable for their operations team.

Looking back at months of operations, George couldn’t help but smile. “The single pane of glass isn’t just marketing. When something breaks, we’re not playing detective across five different dashboards trying to piece together what happened.”

 

Migration: What Actually Worked (and What Didn’t) 

Rather than presenting a sanitized success story, LifeinCloud details the real challenges they faced during implementation:

What Worked Well 

1. Initial Deployment

CloudStack’s basic installation was surprisingly straightforward. George notes: “The core platform was up and running in days, not weeks. The management server installation is well-documented and logical.”

2. KVM Integration

The integration with KVM was seamless. The hypervisor’s mature state and CloudStack’s native support meant we avoided the compatibility issues often seen with newer virtualization technologies.

3. API Architecture

CloudStack’s API-first approach made automation straightforward. Every UI action has a corresponding API call, making it easy to script and automate operations.

Their CTO made a final point about CloudStack’s API: “It’s remarkably stable and well-documented. It hasn’t broken compatibility in years, making it reliable for long-term automation. This is crucial for our customers building infrastructure-as-code pipelines.”

Where They Struggled 

1. High Availability Setup

For LifeinCloud, getting high availability right for the management server was trickier than expected. George explains: “The documentation for advanced HA configurations could be better. We ended up working directly with ShapeBlue and the CloudStack community to resolve some edge cases.”

2. Performance Tuning

While basic setup was easy, optimal performance required careful tuning:

– Ceph required extensive optimization for their workload patterns

– Network configuration needed refinement for maximum throughput

– Management server database required careful sizing and tuning

3. Documentation Gaps

The official documentation, while good for basic scenarios, left some advanced topics underexplored. Community forums and direct interaction with other CloudStack users proved invaluable.

The Business Reality: Beyond Technical Considerations 

The true test of any infrastructure platform is how it supports business objectives. Their CEO, Bogdan, provides perspective: “The real differentiator in the cloud market is economics, not technology. Hyperscalers have created a pricing model that seems simple on the surface but becomes increasingly opaque and expensive as you scale. We wanted to build something different.”

The Economic Advantage 

LifeinCloud’s CloudStack implementation has delivered several key advantages to their organization:

1. Cost Structure

– No per-CPU or per-core licensing fees

– Predictable infrastructure costs

– Lower operational overhead through automation

2. Performance Metrics

– 15% lower latency for intensive workloads

– Improved storage performance through NVMe SSDs integration

– Better resource usage through efficient orchestration

The European Perspective 

Operating as a European cloud provider brings unique challenges, as LifeinCloud’s CEO describes: 

“Europe isn’t one market like the U.S. – it’s many interconnected ones, each with distinct regulations and requirements. A single European hyperscaler can’t address this complexity. What we need is a network of sovereign providers who understand local markets and can move faster than global giants. This is why enterprises are moving away from the ‘hyperscaler or nothing’ mentality.”

Looking Forward: Strategy and Growth 

Market Evolution 

When asked about market trends in the near future, Bogdan, LifeinCloud’s CEO, sees three major ones:

“Choice, speed, compliance – these are the new priorities. Businesses are adopting multi-cloud to stay flexible and avoid vendor lock-in. Users expect near-instant performance, driving the need to deploy compute and storage closer to them. Meanwhile, regulators demand increased transparency and security, especially in Europe. With our choices so far, we feel well positioned to meet all three.”

Technical Roadmap 

LifeinCloud just launched their high-frequency compute offering, S3-compatible object storage, and Kubernetes integration. Upcoming technical initiatives include GPU support with direct device passthrough.

Geographic Expansion 

The cloud solutions provider already has three availability zones and is planning to add one more in 2025.

They chose the Frankfurt availability zone for its proximity to DE-CIX, Europe’s largest Internet peering exchange. Bucharest features some of the fastest broadband internet in the world and is a key hub for Eastern-European markets. The latest one, London, joined their network February – Brexit or not, it’s still Europe’s largest data center market and a global business hub. 

Then came a plot twist, as their CEO explains: “We were all set for Amsterdam, but the restrictions on data centers forced us to look elsewhere. So, by the end of the year, we’re targeting Miami. It connects our European customers directly to the Americas and opens up Latin America.”

George, LifeinCloud’s CTO, adds: “So far, we’re happy with how ACS performed in this expansion: one platform, multiple regions, no silos. We can scale globally while keeping data exactly where it needs to be.”

 

The Real Lessons

Finally, we asked LifeinCloud to give us their top takeaways after running CloudStack in production:

1) Simplicity wins: more features doesn’t always mean ‘better’ and fewer moving parts results in fewer problems.

2) Community matters: open-source is more than code; the real value comes from shared knowledge and collaboration

3) Independence counts: owning your infrastructure means long-term control and flexibility.

In the end, perhaps that’s the true promise of open-source cloud infrastructure – not just the freedom to choose your path, but the power to shape it.