Hi!
I've a Postgres database deployed in my Kubernetes cluster and I'm trying to connect my services with it using Traefik Mesh.
I have the following Postgres services:
---
apiVersion: v1
kind: Service
metadata:
name: postgresql-headless
namespace: postgresql
labels:
app.kubernetes.io/name: postgresql
app.kubernetes.io/version: 1.0.0
app.kubernetes.io/component: infrastructure
app.kubernetes.io/part-of: databases
service.alpha.kubernetes.io/tolerate-unready-endpoints: "true"
spec:
type: ClusterIP
clusterIP: None
publishNotReadyAddresses: true
ports:
- name: tcp-postgresql
port: 5432
targetPort: tcp-postgresql
selector:
app.kubernetes.io/name: postgresql
---
# Source: postgresql/templates/svc.yaml
apiVersion: v1
kind: Service
metadata:
name: postgresql
namespace: postgresql
labels:
app.kubernetes.io/name: postgresql
app.kubernetes.io/version: 1.0.0
app.kubernetes.io/component: infrastructure
app.kubernetes.io/part-of: databases
spec:
type: ClusterIP
ports:
- name: tcp-postgresql
port: 5432
targetPort: tcp-postgresql
selector:
app.kubernetes.io/name: postgresql
role: master
And the following pod deploying (a custom application that I have):
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: business-app
labels:
app.kubernetes.io/name: business-app
app.kubernetes.io/version: 1.0.0
app.kubernetes.io/component: business
app.kubernetes.io/part-of: application-layer
spec:
replicas: 1
selector:
matchLabels:
app.kubernetes.io/name: business-app
template:
metadata:
labels:
app.kubernetes.io/name: business-app
app.kubernetes.io/version: 1.0.0
app.kubernetes.io/component: business
app.kubernetes.io/part-of: application-layer
spec:
containers:
- name: business-app
image: 904333181156.dkr.ecr.sa-east-1.amazonaws.com/business:$ECR_TAG
env:
- name: TYPEORM_HOST
value: '172.20.139.201'
...
ports:
- containerPort: 80
If I use the cluster IP of Postgres service (like above) it works as expected, but as soon as I change it to postgresql.posgresql.traefik.mesh it simply stops working.
The strange thing is that I can curl to Postgres service inside my pod, but here is the log I receive at Postgres's pod:
2020-12-10 23:31:05.093 GMT [7600] FATAL: no PostgreSQL user name specified in startup packet
Is this a bug? Or am I doing something wrong?