What would you like to be added:
Kubernetes Service should support configuring weights for different backend server groups/target groups under the same Service, in order to control the traffic distribution ratio.
Why is this needed:
In production, there are many scenarios that require group-based traffic control, but Kubernetes Service lacks a standardized way to express weights. As a result, teams have to rely on vendor-specific features, manual configuration, or additional traffic gateways, which increases complexity and operational costs.
Typical scenarios include:
Uneven capacity across multiple availability zones (AZs): Some AZs may be capacity-constrained or more expensive, so it’s necessary to reduce the share of traffic sent to those AZs.
Disaster recovery / emergency degradation: When a backend group experiences failures or performance degradation, operators need to quickly lower its weight rather than fully removing it.
What would you like to be added:
Kubernetes Service should support configuring weights for different backend server groups/target groups under the same Service, in order to control the traffic distribution ratio.
Why is this needed:
In production, there are many scenarios that require group-based traffic control, but Kubernetes Service lacks a standardized way to express weights. As a result, teams have to rely on vendor-specific features, manual configuration, or additional traffic gateways, which increases complexity and operational costs.
Typical scenarios include:
Uneven capacity across multiple availability zones (AZs): Some AZs may be capacity-constrained or more expensive, so it’s necessary to reduce the share of traffic sent to those AZs.
Disaster recovery / emergency degradation: When a backend group experiences failures or performance degradation, operators need to quickly lower its weight rather than fully removing it.