Replica override for microservices on VMware
Modify the replica count for microservices by using template overrides for Custom Resources.
When API Connect is deployed, the deployment profile you select drives the number of replicas to
be used for most deployment. For development profiles, typically specified
byn1x....
, the replica count is 1. For production profiles, typically specified by
n3x...
, replica count is 3.
You can override the replica count for individual deployments by using template overrides.
- Template overrides are a piece of yaml that should be added to the Kubernetes custom resource.
The examples provided in this topic show only that the
template:
element should be under the mainspec:
element of a CR. The rest of the CR is not shown. - When applying overrides for Appliance deployments (virtual machines), the yaml sample provided as example must be added in an extra-values-file for the subsystem. To view an example of documenting a template override for an Appliance, see Setting rate limits for public APIs on the management service
Management subsystem
For a Management CR in Kubernetes or OpenShift, the following override specifies 5 replicas for
the apim
deployment:
spec:
template:
- name: apim
replicaCount: 5
juhu
ldap
lur
Analytics subsystem
The Analytics microservices that can be scaled using template override are:
client
ingestion
mq-kafka
mq-zk
mtls-gw
storage-coord
storage-data
storage-master
Note that this microservice cannot be scaled below 2 replicas when using a 3 node (production) deployment profile. Also, when using this
unique
storage type, you must have at least 1 replica available for the 3 storage roles (coord, data and master).storage-shared