The global IT outages that have occurred over the past few years are a clear reminder that service disruptions are inevitable. The ripple effects across banking, retail, and healthcare have shown how quickly operations can be disrupted when organisations rely on a single provider, or even a single cloud region. With this in mind, organisations need to incorporate multi-cloud capabilities into their hybrid cloud strategies.
This is especially significant in the Asia Pacific, where hyperscaler expansion is accelerating at a record pace. According to the Asia Data Centre Landscape report by KPMG, Southeast Asia’s data centre market is projected to more than double, from USD 13.7 billion in 2024 to USD 30.47 billion by 2030 at a compound annual growth rate (CAGR) of over 14 %. The region is witnessing a surge of investments from global cloud and hyperscale operators such as Google, Microsoft, AWS, and Alibaba Cloud, with Malaysia, Indonesia, Thailand, and Vietnam emerging as prime destinations.
With Southeast Asia’s digital economy surging past USD 300 billion, what comes next?
This growing reliance on large-scale infrastructure heightens the risk of single-provider outages, reinforcing the need for hybrid cloud strategies to incorporate multi-cloud capabilities.
Why hybrid must extend to multi-cloud
Hybrid cloud is the keystone for enterprise freedom. It gives organisations the ability to run data and AI workloads where they best support business imperatives. By removing friction and avoiding vendor lock-in, hybrid cloud allows for the seamless movement of workloads across environments. While a hybrid foundation provides crucial freedom and choice, the recent outages reveal a critical lesson: when an architecture simply connects a data centre to a single public cloud provider, the organisation remains dangerously exposed. It has merely traded one single point of failure for another.
This is especially important as hybrid deployments are set to become the new norm. According a recent IDC report, it is forecasted that 75% of enterprise AI workloads in the region will run on hybrid, fit-for-purpose infrastructure by 2027, as organisations seek to optimise performance, cost, and compliance while accelerating time-to-value for AI. As such, organisations need to realise that true scalability and resilience require architectures that avoid single-cloud dependency and support seamless workload portability across environments.
True resilience is about eliminating single points of failure. A modern hybrid strategy, therefore, must be a multi-cloud strategy. Achieving true business continuity means having the freedom to “failover anywhere”, across cloud regions, back to on-premises environments, or between different cloud providers entirely. Business operations must continue, and critical systems must remain up and running, regardless of where the initial disruption occurred.
Additionally, as organisations scale AI workloads, the risks associated with inconsistent governance and security become even more pronounced. Cloudera’s ‘The Evolution of AI: The State of Enterprise AI and Data Architecture’ report indicates that nearly half (48%) of IT leaders in Singapore cite security and compliance risks as a top barrier to scaling AI, with concerns ranging from model manipulation or poisoning (41%) to unauthorized data access (41%) and data leakage during model training (38%).
Multi-cloud resilience mitigates these risks by ensuring that security, governance, and controls remain consistent across environments, even in the face of disruption.
What makes a successful failover strategy
While companies must adopt a multi-cloud failover strategy, it’s incredibly complex in reality. Different cloud providers use different APIs, data services, and security models. For most organisations, moving a mission-critical data workload from one cloud to another requires extensive efforts from application refactoring to redesigning security policies and migrating data. For multi-cloud failover to work, organisations need data and AI services that operate consistently across environments, eliminating infrastructure dependency.
A workload is not defined by data alone; it includes the security controls, governance policies, metadata, and lineage that must move with it. A unified data fabric that maintains consistent metadata, governance, and security across environments is critical in ensuring that workloads can be activated quickly and safely during a failover.
Service disruptions will continue to occur, and the organisations best positioned to withstand them will be those that proactively eliminate single points of failure and design continuity strategies that span multiple clouds and environments.
The article titled “Strengthening hybrid cloud resilience in a world of unavoidable outages” was authored by Remus Lim, Senior Vice President, Asia Pacific & Japan, Cloudera
About the author
Remus is a seasoned enterprise software solution sales leader with more than 25 years of experience under his belt. He is passionate about how data and analytics transform lives, businesses and has been hands-on and instrumental in leading successful teams of change-makers to empower clients on their digital transformation journeys.
Remus first embarked on his data and analytics career with Crystal Decisions/BusinessObjects, followed by leadership roles at IBM’s Information Management and Business Analytics Group, SAP (pioneered the launch of HANA in Southeast Asia) and most recently at Microstrategy Southeast Asia.
Outside of work, Remus’ other passions are centred around his family and kids, keeping a strict fitness routine, pursuing his love for the outdoors and achieving personal challenges such as going on annual hikes within the Himalayan region.