blog: why-lifeincloud-chose-cloudstack

Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>

Author: yadvr

blog/2025-03-18-why-lifeincloud-chose-cloudstack/bogdan.png

@@ -0,0 +1,183 @@
+---
+layout: post
+title: Why LifeinCloud Chose ACS for their Public Cloud
+tags: [case-studies]
+authors: [jamie]
+slug: why-lifeincloud-chose-cloudstack
+---
+
+## The Story of LifeinCloud
+
+[LifeinCloud](https://lifeincloud.com/) is a privately-owned, European public
+cloud provider on a mission to make powerful cloud solutions more accessible to
+developers and businesses worldwide.
+
+They started in 2009 as a managed services provider, later expanding into
+private cloud solutions. As demand grew, they faced a similar challenge to many
+in our community:
+
+<!-- truncate -->
+
+<div className="row">
+<div className="col col--4">
+
+![](bogdan.png "Bogdan Rohan" class="blog-image")
+
+</div>
+<div className="col col--8">
+
+*"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](https://www.linkedin.com/in/bogdanrohan/), LifeinCloud Founder & CEO
+
+</div>
+
+</div>
+
+Their solution? Building a multi-region public cloud platform powered by Apache
+CloudStack. The results transformed their operations: 40% reduction in overhead,
+85%+ infrastructure utilization, and onboarding time cut from weeks to hours.
+
+But the interesting part is how they got there.
+
+## The choice: CloudStack vs OpenStack
+
+During their technical evaluation, LifeinCloud's team discovered significant
+architectural differences between the two platforms. Their CTO, [George
+Lisandru](https://www.linkedin.com/in/george-lisandru-6840a9294/?originalSubdomain=ro),
+explains why OpenStack's architecture proved challenging for what they were
+trying to achieve:
+
+<div className="row">
+<div className="col col--4">
+
+![](george.png "George Lisandru" class="blog-image")
+
+</div>
+<div className="col col--8">
+
+*"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."*
+
+</div>
+</div>
+
+Their testing revealed striking differences in resource requirements for managing 100 VMs:
+- OpenStack needed 5+ control nodes and nearly 100GB of RAM just to manage itself
+- CloudStack ran the same workload with a single management server (two for HA) and just 16GB of RAM
+- Ceph integration with CloudStack "just worked," while OpenStack required additional components and careful configuration
+
+George also singled out CloudStack's API-first approach as a major advantage:
+
+*"It's remarkably stable and well-documented. The API hasn't broken compatibility
+in years, making it reliable for long-term automation. This is crucial for our
+customers building infrastructure-as-code pipelines."*
+
+## Implementation Challenges
+
+While basic deployment was straightforward, the team encountered several
+technical hurdles, as George notes:
+
+"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."
+
+Other challenges included:
+- Extensive Ceph optimization for their workload patterns
+- Network configuration refinement for maximum throughput
+- Management server database tuning for multi-datacenter operations
+
+## Real-World Performance
+
+After optimization, LifeinCloud reports that their deployment is now delivering
+consistent performance:
+
+**Compute Layer:**
+
+- [VM Provisioning](https://lifeincloud.com/products/cloud-servers/): 45 seconds average (down from 3+ minutes)
+- Live Migration: Sub-second downtime for running VMs
+- CPU Overhead: Less than 2% for KVM virtualization
+
+**Storage Layer (Ceph with NVMe):**
+
+- 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
+
+When asked about why Ceph won, LifeinCloud’s CTO explains: "Ceph, while more
+resource-intensive, met our reliability requirements and offered better scaling
+characteristics."
+
+Their 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](https://lifeincloud.com/products/block-storage/) for high-performance workloads
+
+## Network Architecture
+
+Their network design emphasizes redundancy and isolation through:
+
+- Dual physical switches for hardware redundancy
+- Network bonding on all hosts
+- Automatic failover capabilities
+- VLAN-based network separation
+- [Granular firewall](https://lifeincloud.com/products/cloud-firewall/) controls
+- Support for custom network topologies
+
+## Security Architecture
+
+The provider says security was built as a fundamental component into every layer of their stack:
+
+**Network Security:**
+
+- CloudStack's VLAN capabilities maintain complete customer traffic separation
+- Per-VLAN firewall rules allow customers to implement precise security controls
+- No cross-tenant network visibility
+
+**[Access Control](https://lifeincloud.com/features/identity-access-management/):**
+
+- Granular Role-Based Access Control (RBAC) system
+- Precise permission management
+- Comprehensive admin action logging for compliance requirements
+
+*"The centralized management approach is key here,"* notes George. *"When
+something needs investigation, we're not playing detective across five different
+dashboards trying to piece together what happened."*
+
+## Technical Roadmap
+
+The LifeinCloud platform currently operates across multiple datacenters:
+
+- Frankfurt: Chosen for proximity to DE-CIX, Europe's largest Internet peering exchange
+- Bucharest: Features some of the fastest broadband internet globally
+- London: Europe’s largest data center market and a global business hub
+- Miami: Planned for end of 2025, connecting European customers to the Americas
+
+Having just released their
+[Kubernetes](https://lifeincloud.com/products/kubernetes/), [high-frequency
+compute](https://lifeincloud.com/products/high-frequency-compute/), and [S3-compatible
+object storage](https://lifeincloud.com/products/object-storage/) offerings,
+their upcoming technical initiatives include GPU support with direct device
+passthrough.
+
+## The Final Verdict
+
+After running CloudStack in production, LifeinCloud achieved the scalability
+they needed without the complexity they feared. Their infrastructure handled
+growth from one region to multiple availability zones while maintaining
+performance and reliability. Their 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."*
+