| | September 20179CIOReviewThe resilience that ESX provided allowed us to have tremendous success in eliminating most of our Unix based midrange systems and move them to Linux and commodity equipment. The hardware and support costs of these changes saved us millions of dollars in annual expense.VMware also moved out of our datacenters in to our field locations. Cardinal Health has over 600 locations of varying sizes globally. In many of our large locations, we had what I would consider a traditional "server stack" consisting of various management servers, Active Directory controllers, file and print servers etc. In some locations, we also ran local manufacturing or warehouse management systems that interface with our datacenter based SAP instances. In evaluating emerging technologies, we found that if we purchased a small storage device and two computing notes we could develop an easily deployable solution that was extremely resilient and could function on a single node.Cardinal Health currently has over 8,000 virtual servers in our two enterprise datacenters and remote locations globally. As we grow organically and through acquisition, this number continues to rise.Throughout our VMware journey, we have always had to guard against one major failure point. A catastrophic storage failure could have a major impact on hundreds of servers in our datacenters or remote locations. In order to combat that we traditionally spent a lot of time, effort and money on extensive efforts to create backups, storage snap shots and various other High Availability and Disaster Recovery activities. Additionally, we improved our storage monitoring and developed predictive reporting and alerts, but until VMware was able to virtualize the storage layer with VSAN, I knew we had an exposure.Cardinal Health is continuing to move forward in to the Software Defined Data Network by leveraging VMware's virtual storage area network (VSAN) product. Both our virtualization teams and storage teams completed extensive proof of concept work with VMware on various hardware platforms and we intend to move much of our storage to VSAN over the next five years as our current storage is due for refresh. In our warehouses, our new server deployment will no longer have a separate storage device, instead leveraging VSAN. In addition to improving resilience, VSAN will cut the cost of traditional storage by almost 50 percent.I began this article by stating that my primary role is uptime, performance and cost because I believe that anyone working through their own virtualization and software defined datacenter journey should start by ensuring these three things are satisfied. In today's IT environment, though, infrastructure leaders are being asked to do a lot more to enable innovation, improve speed of delivery and provide more self-service to developers and application support teams.Cardinal Health is evolving our internal cloud to improve our capabilities to monitor usage, provide cost transparency and improve efficiency where we see stranded capacities. We have leveraged automation and orchestration by automating server builds through interfacing ServiceNow and VMware's Operations Manager. We are now capable of delivering servers in days through both internal process improvements and automation and we are able to provide temporary space for sandbox and proof of concept work, which has spawned multiple innovative development efforts that previously may have not been realized.Since we started our partnership with VMware and Rolta/Advizex in 2006, both the IT industry and the healthcare industry have changed dramatically. I am impressed that VMware has continued to invest in research and development and acquire the right companies and intellectual property to evolve and improve their product line. We continuously evaluate emerging products but have not seen the pace of improvements and the consistency and reliability in many of the other competitor products that we have seen from VMware. I'm excited to be positioned to continue this journey. We are expanding our hosting options with the use of external cloud providers so I look forward to extending all of our learnings over the last decade to accelerate the new capability, while also looking forward to developing new best-practices that may translate to our datacenters. A catastrophic storage failure could have a major impact on hundreds of servers in our datacenters or remote locations
< Page 8 | Page 10 >