There are two options to scale a component:
- Manually scale the number of replicas
- Configure horizontal auto-scaling
Set number of replicas
To set the number of replicas, simply set the
replicas parameter for a component within the
deployments: - name: my-backend component: replicas: 4 containers: ...
Instead of just running the containers defined in
my-backend with one pod, the above example would run
4 pods with the specified containers. Each of these pods run isolated, i.e. altough the containers of one pod can communicate via
localhost, the containers of different pods cannot.
Configure horizontal auto-scaling
To enable horizontal auto-scaling for a component, you just need to set
autoScaling.horizontal.maxReplicas greater than the value
replicas (see above). Additionally, you should configure one or multiple of the target value parameters,
averageMemory. These target values define how the auto-scaler will set the number of replica to achieve an average CPU utilization and/or an average memory usage by the pods that will be scaled within this component.
deployments: - name: my-backend component: replicas: 4 autoScaling: horizontal: maxReplicas: 10 averageCPU: 800m ...
The above example would create an horizontal po auto-scaler which is configured to:
- create at least 4 pods for the component
- scale the component up to a maximum of 10 pods
- observe the CPU usage of the pods and try to scale between 4 and 10 pods to achieve an average CPU utilization of 800m