German ICT specialist T-Systems, a Deutsche Telekom subsidiary, has been providing Shell sites worldwide with computing power and storage since 2008 – operating from four global data centers in the Netherlands, Malaysia, the US and Germany.
As a cloud computing pioneer, T-Systems is committed to a Zero Outage strategy, implementing a quality assurance program by the same name in 2011. The goal is to minimize downtime and maximize business continuity.
However, Shell's US data center, the Houston Information Center, was built over forty years ago. It was time for Shell to send the data center dinosaur into retirement and, in March 2014, to sign a contract for the Houston Information Center Accelerated Exit Migration program, the largest undertaking of its kind in the company's history.
„The Houston IC Accelerated Exit Migration Program was a very succesful one, especially from a business continuity point of view.“
Doutie Nadema, Manager Shell Supplier Services Platforms and ITSO representative on ERP LT (Run Better)
Over 100 business-critical services, but only one destination
T-Systems’ task was not just to migrate IT resources to the new facility, but also to decommission and dispose of old property. This comprised no less than 69 business-critical IT application landscapes and more than 7,000 IT infrastructure assets (e.g. servers, storage filers, switches and racks) on 3,500 square meters of pure IT space. It was a huge project, so John Kiest, Program Director at T-Systems USA, put together an international team that, over the entire project period, comprised as many as 220 T-Systems professionals.
“Our success was down to careful planning, and close collaboration and communication among all participating teams – including Shell and our technology partners, HP and AT&T," said Kiest. Each and every step was carefully defined, agreed, prepared and executed – with a consistent focus on always making sure someone was in charge, and someone was on duty. This included weekends, when all stakeholders had to maintain sufficient resources available 24 hours a day to be able to send the right experts to intervene if there were complications.
New virtualization technology
Phase 2 was no less strenuous: the business-critical IT landscapes had been successfully migrated. But now the team had to tackle the remaining systems and decommission and dispose of the legacy data center assets in order to allow Shell to put the building on the market for sale. Cisco’s Overlay Transport Virtualization enabled T-Systems to migrate 550 virtual servers for a huge number of dynamic cloud hosted applications and services without having to change their IP addresses.
The move went smoothly. But there was still work to be done. The T-Systems team turned to the hardware that was now surplus to requirements – 7,200 servers, storage devices, tape libraries, server enclosures and other assorted IT assets. They had to be either sold off or disposed of. By late June 2016, this task too was completed, and the legacy data center was handed over on time to Shell’s Real Estate department.
You want to move to another data center? Here are some things you should do:
Identify all the hardware and software that needs to be moved from one data center to the other.
Determine site architecture and design specifications based on the “footprint” required in the new data center.
Complete capex approvals for required new equipment.
Submit procurement requests.
Complete high-level design for the migration architecture and processes based on the estate to be moved.
Commence migration planning based on the high-level design and the agreed-upon migration paths.
Begin staffing up the teams (as required) to do the migration work.
Complete migration wave planning and set it out in writing.
Begin to execute migration waves in accordance with plan.
As waves complete, do the necessary logical and physical decommissioning and disposal.
Wolfram Müller, CIO at Europe’s largest container terminal operator, talks about cybersecurity in the Internet of Things, 99.8 percent IT performance and the advantages of having a “spider in the web”.