Delivering the hands-on support and critical infrastructure that powers Host1Plus
A conversation with Host1Plus CTO Aistis Zenkevicius
At ServerCentral, we’ve been providing the hands-on support and delivering the critical infrastructure that service providers need since our inception in 2000.
Over the past 17+ years we’ve come to learn that there really isn’t anything better than the respect of our peers. This is why we pride ourselves on being the service provider’s service provider. Whether it’s delivering the high-performance infrastructure that powers our partner’s business or a deeper partnership involving the architecture, design, deployment, and management of a custom private cloud, we work extremely hard to understand and match the level of care our service provider partners promise to their own customers.
When global hosting provider Host1Plus approached us, we welcomed the opportunity to share our experience and learn how it could apply to helping them grow. Since 2008 Host1Plus has been offering OpenVZ and KVM virtualization services, with an emphasis on custom networking solutions. Host1Plus serves power users of Linux or Windows-based services — the kind who require APIs, intuitive UIs, and powerful, on-demand resources.
To learn more about Host1Plus and what service providers look for when evaluating other service providers, we sat down for a conversation with Host1Plus CTO Aistis Zenkevicius.
Q: Please introduce yourself.
I’m Aistis Zenkevicius, and I’ve been in technical jobs a tad shy of 20 years, with a keen interest in the Infrastructure as a Service field.
As CTO at Host1Plus, I guide the infrastructure team, sharing my know-how in areas spanning from storage system design and virtualization to system monitoring. I organize the entire development life-cycle, performing day-to-day project management and defining product roadmaps. I work closely with the management team to make sure the product team delivers the capabilities required to execute business goals, and to help with global expansion efforts. I also work with clients on custom solution design.
Q: What was the catalyst that caused Host1Plus to look for a new infrastructure provider?
The most important business challenge we had to solve was having top-notch remote hands manning our critical infrastructure 24 hours a day, 7 days a week. Our customers expect and require our infrastructure to always be available. Having a partner that understood, at its core, what always-on means was critical.
As of this moment, we have yet to find another infrastructure provider with as high a level of professionalism as ServerCentral.
Q: What changed with your previous provider?
ServerCentral was the first data center in the U.S that moved from a pay-per-use model to allowing fully owned rack space, power and compute. Previously we were unable to control which transit providers we could use or how we could grow our rack space. We jumped at the opportunity to be more strategic and forward-thinking.
Q: What technological challenges did you need to solve?
The two biggest were technological challenges are really core infrastructure challenges — power consumption per rack and available connectivity. At Host1Plus, we tend to use high power-density compute solutions that consume the traditional data center rack power cap at half the rack capacity. Obviously, this is not efficient — technically or financially. We were looking for a partner who could deliver the power quality and density that we need. We’re a service provider — we needed a true partner, another service provider capable of meeting the same standards we promise our customers. By having the NOC in-house, by having people on the floor around the clock, by deeply understanding our business, ServerCentral helped us to bring our services up in the shortest possible time.
Q: What convinced you ServerCentral could help you hit these goals?
ServerCentral had a good location, good power price, blended connectivity, and best-in-class remote hands. When you meet an organization that understands the needs of service providers, and is ready, willing, and able to support them, you know you’ve found the right partner.
Working with the ServerCentral team, we were able to get racks with 8kW (or more) of available power. Most importantly, we have real-time visibility into our power consumption per rack in the ServerCentral portal. We’re then able to connect this information with our own internal system monitoring and control compute node power consumption by tuning CPU performance. ServerCentral also gave us the option to merge our multi-rack power commit into a single pool when we needed, which greatly simplified our footprint management and expansion.
In addition to allowing us to control our usage and costs, ServerCentral provides us with blended, redundant traffic and redundant links that deliver a true 100 percent SLA on connectivity.
Q: Tell us about the migration process.
Well, we got more than we expected. Kudos to ServerCentral! The ServerCentral team worked with us and took our needs to heart as if they were part of Host1Plus. They proactively created a rack diagram, and they performed the installation without a hiccup. The team at ServerCentral has also saved us a lot of time by providing the required network and power cabling.
We’ve been pleasantly surprised by how quickly and efficiently ServerCentral staff addressed our requests and concerns. Going forward, I believe the only real challenge will be to continue to correctly forecast the growth and agree on floor space.
Q: What does the next year hold for Host1Plus?
As mundane as it sounds, our goal is to capitalize on our cloud product and crystalize our client base. Global political instability, currency fluctuations, and hardware component shortages (e.g. RAM modules) all put stress on our business. They add unpredictability, and unpredictability leads to indecisiveness in the market. The continuous price damping by the big three IaaS providers doesn’t help much, either. We will continue to remain focused our fundamentals to help us ride out these waves.
To help differentiate Host1Plus in the crowded hosting area, we are working on new and unique networking products. ServerCentral’s geographical location, as well as having available transit providers on site, will help us to achieve our goals.
Q: What changes can we expect in the broader enterprise IT industry over the next 18 months?
Over the next 18 months, we should see a lot of things changing in the enterprise IT industry. On the storage side, we will see solid-state disks become the de-facto standard. Interfaces, like SATA, will surely be displaced by high speed PCIe/NVMe/M.2 on most of the servers, except for backup and long-term storage machines. As the semiconductor yields improve, we will see price drops and increased endurance products become available at reasonable prices. The trend to build APIs and use them to connect applications and/or services will continue. People will continue to consume more network bandwidth and storage space as content quality continues to increase.
Expect providers, enterprises, and businesses to move beyond 40Gbps link speeds and into 100Gbps over the next year and a half. We are also looking forward to reduced power consumption across the board. As spinning disks get replaced by solid state and BASE-T connectivity is replaced by Fiber, as hyper-efficient CPUs and memory modules (via improved lithography processes) all come to market, the price-to-performance ratio is going to change dramatically. Seeing how companies adopt these new capabilities, and what they build on top of it, is exciting.