Cassava Technologies, through its Africa Data Centres unit, has brought a long-awaited resilience upgrade to South African businesses: its JHB1 data center in Johannesburg is now a Microsoft Azure ExpressRoute Metro peering location. The designation, announced on July 20, makes Johannesburg the first metro area in Africa where enterprises can split a single ExpressRoute circuit across two physically distinct sites—effectively insulating private cloud connections from a building-level outage.
For the thousands of organisations already relying on Azure from Johannesburg data centers, the move addresses a persistent single point of failure. Instead of having redundant connections terminate at one facility, the Metro design spreads those links across two locations, so a fire, power failure, or fiber cut at one site doesn’t automatically sever the corporate lifeline to Microsoft’s cloud.
A Second Peering Point Ends the Single-Site Gamble
Until now, a conventional ExpressRoute circuit in Johannesburg provided built-in redundancy through dual connections, but those connections terminated at the same physical peering location. If that facility went dark, so did the enterprise’s private path to Azure. The new Metro configuration changes the equation: JHB1 becomes the second peering endpoint in the city, joining an existing location that had served as the sole on-ramp.
Microsoft’s ExpressRoute Metro isn’t a new concept globally—it already exists in select cities in the Americas, Asia, and Europe—but its arrival in Africa reflects a maturing digital infrastructure. By design, a single Metro circuit bridges two peering sites within the same metropolitan area. If one site suffers an isolation event, traffic can fail over to the other, maintaining connectivity for mission-critical applications.
Cassava is uniquely positioned to capitalize on this. Its subsidiary Liquid C2 has on-net presence at both Johannesburg endpoints, and it plans to package the dual-site connectivity into a managed service called Secure CloudConnect. The offering bundles Azure private connectivity with integrated cybersecurity, sparing customers from having to stitch together carrier, colocation, and security components on their own.
“Being designated a Microsoft Azure ExpressRoute Metro peering location is a defining moment for Africa Data Centres as well as for Africa’s digital infrastructure,” said Ziaad Suleman, Senior Vice President of Cassava Technologies South Africa and Botswana. “It reflects growing confidence in the continent’s ability to support the next generation of cloud and AI-driven services.”
Managed Resilience: Liquid C2’s Play
For a company grappling with budget and skills constraints, Liquid C2’s Secure CloudConnect could lower the barrier to adoption. The fully managed service promises to deliver resilient private connectivity and layered security in one contract. That’s a compelling proposition for regulated industries—banking, insurance, healthcare—where compliance mandates robust business continuity and data protection.
Cassava claims it is the only provider with an on-net presence in both peering locations, which means it can orchestrate the entire data path from the customer edge to Azure. Others may eventually offer connectivity through one or both sites, but for now, the managed route runs through Liquid C2.
The announcement comes as regulators across Africa tighten operational resilience requirements. The South African Reserve Bank, for instance, expects financial institutions to demonstrate credible recovery strategies. Having site-diverse Azure connectivity doesn’t satisfy all regulatory demands, but it removes a glaring infrastructure risk that auditors often flag.
For IT Teams: More Resilient, But Not Infallible
While the Metro designation raises the bar, IT administrators must understand its limits. Microsoft’s documentation distinguishes between standard resilience (redundant connections at one site), high resilience (Metro—two sites, one circuit), and maximum resilience (separate circuits across different metropolitan areas). Metro protects against a localized incident; it doesn’t guard against a city-wide disaster or a wide-area network failure.
Moreover, the weakest link might still be inside the customer’s own environment. If an organization’s edge routers, firewalls, or WAN circuits aren’t configured for seamless failover, the benefits of a dual-site ExpressRoute can evaporate. Admins need to verify that their Azure ExpressRoute gateway SKU supports Metro, that routing protocols are properly set up, and that failover actually works as expected under real-world conditions.
“Customers shouldn’t assume their existing circuit automatically gains Metro resilience,” said one South African cloud architect familiar with the setup. “They’ll need to check with their provider, possibly reprovision the circuit, and definitely test the failure scenarios.”
For hybrid workloads—Windows Server, Azure Stack HCI, Azure Virtual Desktop—the connectivity upgrade could be transformative. Imagine a manufacturing plant running local Azure Stack HCI for latency-sensitive operations, with backup and management traffic flowing to Azure. A building outage at the primary peering site would have previously meant complete loss of cloud synchronization. Now, a second path keeps those sync jobs alive.
The Long Road to Resilient Cloud in Africa
Microsoft turned on its first African cloud regions—South Africa North (Johannesburg) and South Africa West (Cape Town)—in 2019. Those regions brought Azure infrastructure closer to local users and satisfied data residency requirements. But peering locations, which are physical meet-me rooms where enterprise networks connect to Microsoft’s edge, remained a single-site affair in each city for ExpressRoute. That never felt like true resilience for companies that had poured resources into Azure adoption.
The gap has become more acute as African organisations accelerate cloud migrations and deploy AI workloads. A 2024 survey by a Johannesburg-based analyst firm found that 67% of large South African enterprises cited connectivity resilience as a top concern when moving critical systems to the cloud. The same study noted that nearly a third had experienced at least one major Azure connectivity incident in the previous two years.
Globally, Microsoft has been expanding ExpressRoute Metro to more cities, but the pace in Africa has lagged. Johannesburg’s designation now aligns it with peers like São Paulo, Singapore, and Stockholm, where dual-site ExpressRoute is already available. The listing on Microsoft’s official peering locations page confirms that Johannesburg now supports Metro, though the documentation doesn’t call out the specific provider—Cassava’s announcement effectively fills that gap.
Your Migration Checklist
For organizations that want to move from a single-site ExpressRoute to the Metro configuration, the path forward is relatively straightforward but not trivial. Here’s a step-by-step guide based on Microsoft’s guidance and early adopter feedback:
- Inventory your current setup. Identify the existing ExpressRoute circuit, the peering location it uses, and the Azure gateway type. Note the bandwidth tier and whether you’re using standard or premium add-ons.
- Check provider support. If you’re not using Liquid C2, contact your connectivity partner to confirm whether they can deliver ExpressRoute at both Johannesburg locations. Not all carriers have on-ramps at both sites.
- Evaluate the gateway SKU. Metro requires an UltraPerformance or ErGw3Az SKU in some scenarios. Review Microsoft’s ExpressRoute gateway documentation to ensure your gateway can handle the dual-site design.
- Reprovision the circuit (if needed). You may need to delete the existing circuit and create a new one that references both peering locations, or you might be able to modify an existing circuit with provider assistance. Expect a brief outage during the cutover.
- Test, test, test. Simulate a failure at one peering site by withdrawing BGP routes or physically disconnecting. Verify that traffic fails over within the expected time—typically a few seconds—and that applications remain responsive.
- Harden your WAN and edge. Ensure your branch offices or on-premises data centers have redundant last-mile connections and that internal routing isn’t locked to a single path.
- Consider the managed option. If you lack in-house expertise, evaluate Liquid C2’s Secure CloudConnect. It may simplify operations and accelerate time-to-resilience, though it adds a recurring managed service cost.
For budget-conscious teams, it’s worth noting that ExpressRoute Metro isn’t free. There may be additional port fees, circuit charges, and potential hardware upgrades. But for organizations running Tier-1 workloads—especially those in financial services, e-commerce, or healthcare—the improved uptime often justifies the expense.
What’s Next
Cassava’s move is unlikely to be the last. As more hyperscale clouds land in Africa and regulators raise the bar on operational resilience, expect ExpressRoute Metro to appear in Cape Town and perhaps in Nairobi, Lagos, or Cairo as well. Liquid C2 has hinted that it will leverage its pan-African fiber network to extend the Secure CloudConnect model to other metros once Microsoft opens additional peering locations.
The Johannesburg designation also signals a broader maturation of African cloud infrastructure. With AI workloads demanding sustained high-throughput connections, the old single-site ExpressRoute model looks increasingly fragile. Enterprises that treat resilience as a core design principle, not an afterthought, will find that the Metro upgrade aligns perfectly with their long-term hybrid and multi-cloud strategies.