By Dan Newton, CEO, CDS
There is a renewed reliance on service providers to augment, or even replace, in-house infrastructure support for data centers. Consolidation and public cloud adoption are changing how organizations are approaching their infrastructure strategies. In fact, according to a 2020 Gartner research report, it’s expected that by 2025 up to 85% of infrastructure strategies will integrate on-premises, colocation, cloud, and edge delivery options. This is up significantly from the 2020 numbers.
Over my career, I’ve spent (many) years managing service provider data centers -- from Datapipe to Rackspace, I’m aware of all of the challenges associated with maintaining storage, server, and networking hardware for extended periods (and ensuring that everything is done profitably). While each type of equipment is unique in terms of its reliability and serviceability, many of the principles remain the same for building a multi-year strategy that balances profit and SLA delivery.
Service providers are as unique as the equipment they service. Some provide Infrastructure-as-a-Service (IaaS) with hardware stacks designed to support a wide range of customer workloads. Others provide co-location services, hosting customer-owned equipment within managed racks and/or cages.
IaaS providers tend to be relatively homogenous, with minor variations based on single-vs-multi-tenancy, performance requirements, and capacity. The goal of these service providers - just like any capital project - is to maximize the occupancy of the infrastructure for as long as possible - which means operating well past the normal warranty period. Looking at the economics of this model, about a third of the cost is typically hardware, with only 10% going toward maintenance. The largest cost item in this model? Labor, clocking in at almost 40% of the overall costs to manage the environments.
Co-location, on the other hand, provides more flexibility with SLAs varying by service provider and contract. The nuances of systems maintenance - especially for mature assets - are an important point of the negotiations, and often result in lower costs.
It is extremely important to remember that most system updates occur within a device’s original warranty period. Firmware updates, service packs, security patches, and other routine maintenance occur most frequently during the first two years after a release – and then tail off considerably in year three, depending on the system type and vendor. After six years these updates are either infrequent or nearly non-existent.
How does this affect your relationship with a vendor? Or your options? Well, most enterprise OEMs have a variety of pick and choose maintenance options, from premium, high-cost options that deliver white-glove service, to cost-effect support solutions that are better suited to mature systems, especially those that are nearing the end of service life.
Organizations needing to make budget-conscious and quick decisions as they consider their options for the post-warranty support of their data center server, network, and storage systems, choosing between IaaS and Co-location is critical. There is another solution, however. Multi-Vendor Services (MVS) – popularized by many of the larger infrastructure OEMs – to the mix of terms. Multi-Vendor Services (MVS) give your company the support services you need to keep mission-critical components running and significantly reduce cost, partnering with you to support a full stack of capabilities for a wide range of different OEM hardware platforms, from midrange to high-end storage and leading server and networking products.
Modernization initiatives have made the role of data center service providers more critical than ever. There is pressure coming from all angles: infrastructure, talent acquisition, application time to market, etc. Luckily, some providers can help you deliver quality vendor-authorized solutions at a low cost.
About The Author
Dan is the CEO of CDS. Previously, he was at HiveIO, a developer of intelligent virtualization technology, where he served as CEO. Prior to HiveIO, Dan was Chief Operating Officer for the managed services company Datapipe, supporting customers globally in over 20 data centers and across private and public cloud solutions. Datapipe was acquired by Rackspace, where Dan served as GM and SVP of Service Delivery. Dan also has held key leadership roles with Dell and Perot Systems.