Migrating OSS/BSS to the Public Cloud
James Crawshaw, Senior Analyst – Service Provider IT and Automation, Heavy Reading
CIOs are a cautious bunch. Their average tenure in a company is just 4.3 years, according to a survey by Korn Ferry. That is around half the tenure of the CEOs who fire them when IT projects go wrong.
So, understandably, they are nervous about moving applications to the public cloud, outside of their control. What if the data is compromised? What if the public cloud provider has an outage? Sure, public cloud might be fine for consumer services like Netflix; it might be okay for development work or public facing websites. But is a telco CIO really going to run mission-critical production apps, that form part of the Operations and Business Support Systems suite, on the public cloud?
According to a 2017 TM Forum CTIO survey, the majority of CSPs had less than 10% of their BSS in the cloud (private or public). That has probably increased somewhat during the last two years, but, overall, OSS/BSS is still primarily hosted on physical servers and in a telco's own data center. The trend toward cloud hosting is clear, but are CIOs going to migrate OSS/BSS to public cloud given that, on average, they only have another two years left in the job? "Maybe that migration is better left to my replacement," they might think.
Today, the use of public cloud to host OSS and BSS is typically restricted to new entrants, including new brands that existing MNOs launch for market segmentation purposes. Take, for example, Verizon's prepaid, bring-your-own-device brand Visible, which is run and managed completely from public cloud platforms.
Visible's CIO initially considered re-using Verizon's IT stack but soon realized this would not be agile enough to react to changes in the marketplace. Going cloud first also enables Visible to scale its IT resources as the business grows. As Visible's Head of Technology, Adil Belihomji, states, "Scaling with our customers makes much more sense than just rolling in millions and millions of dollars spent on hardware." The net result is a fraction of the cost of the legacy OSS/BSS solution.
Established operators are also starting to wake up to the cost saving and agility benefits of public cloud for their core business. For example, UK mobile operator Three is in the final stages of deploying multiple applications, including its OSS and BSS, on public cloud platforms. This will reduce the number of servers the company operates from 2,400 to just 40 and move more than 90% of its workloads into the cloud. Overall IT costs are expected to drop by 30% compared to the old mode of operations.
Vodafone also sees the public cloud as part of its OSS evolution strategy. Earlier this year the company selected a service assurance solution that was delivered as-a-service, hosted on AWS. More recently, AT&T announced it would move the majority of its non-network applications (i.e. OSS and BSS) onto Microsoft Azure.
So, what is holding other operators back? Privacy and security concerns, performance reliability, lack of skills, and general resistance to change are some of the main factors. These must be overcome in order to get the potential cost savings and agility of public cloud. The best software developers want to work on public cloud, so if CSPs want to recruit and retain them, they need to align their OSS/BSS architecture accordingly.
At Light Reading's Software-Driven Operations summit on November 5 in London, one of the panel discussions will discuss the use of public cloud to host OSS/BSS. We'll examine the challenges of running real-time OSS/BSS applications in public cloud infrastructure, and how these assets can be leveraged to complement private cloud resources and help the migration away from legacy systems. I hope you can join us there.
— James Crawshaw, Senior Analyst, Heavy Reading