3. Orchestration (Infrastructure Scaling) : 3.1 Scaling Localization
3.1 Scaling Localization
When scaling our dynamic infrastructure, we need to know where to scale. Depending on the complexity and architecture of the application service, there are several approaches to scaling.
*Fixed scalingScale where the bottlenecks typically occur.
*Scale everythingScale out the entire environment with each scale event.
*Intelligent scalingScale where the resources are needed.
Table 2. Scaling Modality Benefits and Drawbacks
Scaling Mode
Fixed Scaling.
*Can create bottlenecks in other areas of the service.
*Bottlenecks not within the fixed scaling components are not addressed.
*Scaling might not address the problem, and if not managed properly, this can result in a runaway scaling event.
Scale Everything.
*Can result in overprovisioning in certain tiers of the service.
*Can be far more time consuming during the orchestration phase of scaling out.
*Can be more complicated than a fixed scaling approach due to database configuration and synchronization challenges.
Intelligent Scaling.
*Adds capacity where it is needed every time.
*Scaling out across tiers and components dynamically avoids creating new bottlenecks.
*Excess capacity is not added where it is not required.
In the context of scaling, the scale remediation can scale out, scale up, or a combination of both depending on the level of complexity within the system. Make scaling design decisions based on prior knowledge of the application or services scaling characteristics.