From bare metal to VMware
A storage project that could not meet its original latency requirements became a successful virtualization platform—and a turning point in how I thought about infrastructure.
> €10,000 monthly rack-cost savingsRead the project story
A project I am especially proud of
As a systems administrator, I could change configurations and architecture, but I could not patch away every hardware limitation. We had bought a storage system to provide a cache layer between the web servers and the production systems.
The proof of concept was not the whole picture
The system passed our proof of concept. Under real operational load, however, its latency was too high for the intended role. A project with an investment of several hundred thousand euros was at risk. Did we have to accept failure and write off that investment?
A different use for the same investment
We changed direction. Rather than abandon the storage, we used it as the foundation for virtualizing our bare-metal servers with VMware. The original cache-layer plan did not work out, but the hardware still had value in a different architecture.
A setback became a measurable success
The result was a successful virtualization project. Consolidating the physical servers also saved us more than €10,000 per month in rack costs. We had turned an investment at risk into a platform with a clear operational and financial benefit.
Why this led me toward AWS
Even though the outcome was a success, the experience left me wanting more: infrastructure I could create and change dynamically, much like software. That desire stayed with me. It helps explain why I eventually lost my heart to AWS.


cloud / systems