The Met Office is upgrading its mainframe, with the help of IBM and Computacenter, allowing the national weather service to process greater volumes of weather data more rapidly.
The organisation has operated since 1854 and accounts for as much as a fifth of all government web traffic on a peak day.
It is also undergoing a significant IT transformation that has seen it adopt some cloud services and an API-led development strategy.
“Every day, our weather and climate forecasts help people make better decisions. We help organisations save lives, reduce costs, meet schedules and encourage growth,” said Richard Bevan, head of IT infrastructure and operations at the Met Office.
“We want to transform weather intelligence into a tool that drives greater business performance in today’s digital world and the mainframe is at the heart of that.”
“Any issues with performance can have a direct impact on the availability of the services we provide,” added Martyn Hunt, technical lead for the Met Office’s mainframe team. “Without accurate and timely weather information, everything from commercial flight schedules to emergency rescue operations can be disrupted.”
The new environment, which includes two mainframes with a total of 44 cores and 200 terabytes of attached storage will be able to perform more than 23,000 trillion calculations per second. The platform is based on two IBM z13 LinuxOne mainframes and IBM hybrid storage systems, and will eventually include IBM Wave for virtual server management.
The Met Office says it plans to migrate several additional applications, open source databases and workloads to the new environment, including its file transfer hubs, which are currently hosted on disparate server clusters.
The new mainframe will help power a range of public forecasting services, commercial products and mobile apps at a time of increased demand for weather intelligence. With more than 35 percent additional mainframe capacity, as well as increasing operational agility and efficiency, the upgrade has also reduced total cost of ownership.