Cloud applications and distributed data centers can make enterprise bandwidth requirements difficult to predict. A connection sized for normal operations may be insufficient during a workload surge, while a permanently oversized service can leave capacity underused. G.HAO ODUflex offers a way to adjust transport capacity as requirements change, without interrupting traffic on the connection.
For service providers, the combination of OTN switching and hitless bandwidth adjustment can support more flexible Layer 1 private-line and data center interconnect (DCI) services. The practical opportunity is to align capacity more closely with demand while retaining the traffic separation and service characteristics associated with optical transport.
How G.HAO ODUflex works
ODUflex is an optical transport network (OTN) container that can be sized to carry packet-based services at rates suited to their requirements. In the G.709 framework described in the whitepaper, ODUflex (GFP) capacity can be set from 1.25 Gbps to 100 Gbps in 1.25 Gbps increments. Packet traffic can be mapped into a suitably sized container and carried over higher-order OTN capacity.
G.HAO—hitless adjustment of ODUflex (GFP)—is the mechanism for changing that container’s rate while maintaining the end-to-end connection. Defined by ITU-T Recommendation G.7044, it coordinates the adjustment across the connection path so existing traffic can continue during the change. This distinguishes G.HAO from simply provisioning a new service or taking a connection down to alter its capacity.
In a switched OTN network, sub-wavelength connections can be groomed onto higher-capacity transport paths. Intermediate nodes participate in carrying and adjusting the connection. When capacity needs to increase, the network must have suitable resources along the route—or a route with available capacity—to accommodate the change.
Why scalable Layer 1 private lines matter
Enterprise cloud adoption changes both where data is processed and how much capacity is needed between offices, data centers, and cloud service locations. Some workloads generate variable traffic, while others—such as replication, backup, or large compute tasks—may create distinct periods of high demand. A fixed-rate service can be a poor fit when the required bandwidth changes over time.
Traditional Ethernet private-line and virtual private-line offerings may not meet every requirement for mission-critical cloud workloads. The whitepaper identifies traffic isolation, stringent service-level agreements, and capacity beyond the rates commonly available in those services as important considerations. OTN-based Layer 1 transport can provide a different service model, with ODUflex allowing the connection rate to be matched to demand.
For example, an enterprise might use a private line for site-to-site traffic and a separate connection to a data center. If a new cloud workload increases the data-center link’s requirement, a G.HAO-enabled service could scale that connection rather than requiring the customer to keep a much larger fixed-rate service provisioned at all times. The capacity change is intended to occur without disrupting traffic already on the path.
Using network capacity more flexibly
G.HAO may also help operators make better use of capacity already deployed across metro and core networks. OTN switching can groom traffic and direct connections through available paths in a mesh network. When demand changes, service providers can draw on capacity available elsewhere in the network, subject to the resources and routing options available on the connection path.
This approach is different from assuming that any bandwidth can be made available instantly. A hitless rate adjustment depends on coordinated support across source, intermediate, and destination nodes, as well as sufficient transport capacity. Network planning, service orchestration, and the availability of compatible equipment remain important to the service design.
Time-based DCI and the “follow the moon” model
Data center operators may schedule bandwidth-intensive tasks for periods when demand from other services is lower. The whitepaper describes a “follow the moon” model in which a provider could make more capacity available for DCI workloads during off-peak hours and reduce the service rate during busier periods. This can create a way to match network resources to changing time-of-day requirements.
Such a service would require the operator to coordinate customer demand with capacity across the OTN network. It may also allow existing infrastructure to support an additional service use case, rather than relying only on building new capacity. The benefit depends on the actual traffic patterns, available headroom, and operational arrangements between the provider and customer.
What G.HAO-capable systems need
Hitless adjustment is a coordinated network process, not just a setting on an individual interface. The whitepaper highlights several requirements for OTN processing and system design:
- Accurate rate tracking: Source, destination, and intermediate nodes must follow the changing ODUflex rate.
- Coordinated transitions: Nodes along the path must monitor and coordinate changes so bandwidth is added or removed in step.
- Flexible OTN multiplexing: Multiplexing, demultiplexing, and framing resources must accommodate the revised connection capacity within the transport structure.
- Traffic integrity: The adjustment process must preserve existing traffic and avoid compromising other services carried by the system.
These requirements make interoperability and software support important considerations, alongside the capabilities of OTN processing silicon. The connection spans equipment and nodes, so a service provider needs confidence that the relevant elements can coordinate the adjustment end to end.
Practical implications for service providers and customers
For service providers, G.HAO ODUflex can support service offerings that scale capacity without treating every demand increase as a permanent upgrade. For enterprise customers, it can provide a path to request transport capacity that better reflects workload needs, while retaining a Layer 1 service model.
Successful deployment still requires careful assessment of network capacity, route diversity, node support, service-level objectives, and the operational process for requesting rate changes. For technical decision-makers, the key question is not only whether equipment supports G.HAO, but whether the full connection path can deliver the intended service consistently.
In short, G.HAO combines flexible ODUflex sizing with coordinated, hitless adjustment across an OTN connection. When supported end to end and backed by available network capacity, it gives operators a technical foundation for scalable private lines and time-variable DCI services—and gives customers a way to align transport bandwidth more closely with changing demand.