IT operations and service assurance for four brands

Telia Norway operated four brands on an estate that had grown by acquisition and by project: separate data-centre footprints, separate infrastructure, and operational practices that reflected where each brand had come from rather than what the group now needed.

Head of IT Operations & Service Assurance from 2014 to 2018 meant owning that estate end to end — the infrastructure, the operational processes and the assurance of the services that ran on them — through the period in which Telia's first observability platform and first hybrid cloud were introduced.

Raise the operational efficiency of four brands as one operation, consolidate infrastructure that had never been designed together, and migrate data centres — without an outage window that four brands' customers would tolerate.

• Four brands with different infrastructure, tooling and operating habits.
• Data-centre migrations with live services on both sides of every move.
• Consolidation that had to remove cost without removing resilience.
• An assurance function that needed to see services, not just servers.

The approach treated the four brands as one operated estate with brand-specific content. Infrastructure was consolidated onto shared platforms; data-centre migrations were sequenced by service so that each move could be validated before the next; and service assurance was built above the infrastructure so that the operational view was of services and their dependencies.

Head of IT Operations & Service Assurance: overall responsibility for the operational efficiency of the four brands, including the data-centre and hybrid-cloud infrastructure, and for delivering the infrastructure consolidation and data-centre migrations.

Operations moved from brand-by-brand to one function with one view. The consolidation and migrations changed where and how services ran; the observability platform and hybrid cloud delivered in the same period changed what the operation could see and how fast it could add capacity.

Infrastructure consolidated and data centres migrated across the four brands, with the operation running throughout. Efficiency and cost figures are not published here.

• Migrate by service, not by rack; every step should be reversible until it is validated.
• Consolidation that removes resilience is a deferred outage.
• Assurance has to be designed into the operation, not bolted on after the migrations.