kubernetes YML, use for loop - kubernetes

Question

Is there a mechanism in a template (yml) file in k8s for generating multiple items? Like a for loop.
For example , I have a template used my multiple projects. But some need 1 database others maybe 2, 3 or more. I don't want to define all variables by hands, and edit my file if a new projects comes in and needs a new variable.
Example of what I do today:
template:
(...)
containers:
(...)
env:
- name: DATABASE_NAME
value: "{{prj.database.name}}"
- name: DATABASE_NAME_2
value: "{{prj.database.name.second}}"
- name: DATABASE_USER
valueFrom:
secretKeyRef:
name: "{{ k8s.deploy.app.secret.name }}"
key: database_user
- name: DATABASE_USER_2
valueFrom:
secretKeyRef:
name: "{{ k8s.deploy.app.secret.name }}"
key: database_user_2
As you can see I have to copy paste code for each user + password + database name. In this file and also in secrets. I use XLDeploy to insert data in my YML files.
What i'm looking for:
template:
(...)
containers:
(...)
env:
---- for i in {{prj.database.number}} ----
- name: DATABASE_NAME_{i}
value: "{{prj.database.name.{i]}}"
- name: DATABASE_USER_{i}
valueFrom:
secretKeyRef:
name: "{{ k8s.deploy.app.secret.name }}"
key: database_user_{i}
---- end for loop ----
So I could insert in XLDeploy my number of databases to use for each projects. And fill the values with XLD too.
Is that possible, or I need to use a script langague to generate a YML file from a "YML template" ?

Answer 1

You could use helm for this if you have other yamls related to this file. Helm allows for syntax like this:
{{- range .Values.testvalues }}
---
apiVersion: v1
kind: Service
metadata:
annotations: {}
labels: {}
name: {{ . }}
spec:
ports:
- name: tcp-80-8080
port: 80
protocol: TCP
targetPort: 8080
selector:
app: rss-{{ . }}
type: ClusterIP
{{- end }}
So with a list for testvalues in the values.yaml like this:
testvalues:
- a
- b
- c
- d
Helm would generate four service declarations -- one for each item. The first looking like this:
---
apiVersion: v1
kind: Service
metadata:
annotations: {}
labels: {}
name: a
spec:
ports:
- name: tcp-80-8080
port: 80
protocol: TCP
targetPort: 8080
selector:
app: rss-a
type: ClusterIP
The other option if it's just a single file, is to use Jinja and something like Python to build the file using the jinja2 module