Definition
Geographical Scalability ( Distributed System)
Geographical scalability is the dimension of scalability concerned with operating across greater distances and more widely dispersed network topologies while maintaining the required quality of service.
The system must accommodate increased communication delays and differing bandwidth constraints. It cannot rely on every interaction having the low latency of a local area network; geographical scalability does not remove the physical cost of distance.
LAN and WAN Assumptions
Moving from a LAN to a WAN changes which communication assumptions are practical. These are common design tendencies, not guarantees implied by geographical scope.
| Aspect | LAN | WAN |
|---|---|---|
| Timing | A synchronous model with known timing bounds may be practical in a controlled deployment. | An asynchronous model is often safer: correctness does not rely on a known upper bound on message delay. |
| Speed | Round-trip latency is typically low. | Greater distances and more network hops increase latency. |
| Discovery | Local broadcast may locate resources within a subnet. | Communication is usually directed to known endpoints; network-wide broadcast is generally unavailable. |
| Reliability | Failures may be less frequent and easier to control within one site. | More links and administrative domains introduce additional opportunities for loss, disconnection and network partitions. |
Here, synchronous and asynchronous refer to timing assumptions, not whether an application makes blocking calls. A LAN does not automatically provide timing bounds or reliable delivery, and a WAN does not rule out explicitly engineered guarantees.
Consequences for Design
- Reduce sequential exchanges: an operation requiring
request–reply rounds incurs roughly communication delay when each round has latency , even before processing time is counted. Several cheap LAN exchanges can become expensive across sites. - Avoid blocking unrelated work: asynchronous requests allow useful work to continue while waiting for remote responses. They do not make those responses arrive sooner.
- Replace broadcast discovery: a location service that finds resources by asking every local machine must use another approach across a WAN, such as a directory, hierarchical lookup or cached location information.
- Handle communication failures explicitly: timeouts, recovery and retries must account for delayed or lost messages. A timeout alone cannot distinguish a failed participant from a slow or unreachable one.