On Debian 11, I'm trying to do a deployment with Kubernetes from one I did with Docker compose until now, to install a PostGIS instance.
My cluster is a Minikube with CRIO. And I essentially followed what I found over the Internet.
Eventually, everything look running, but a connection test to the pod with psql fails.
Here are my deployers descriptors:
- Namespace:
apiVersion: v1
kind: Namespace
metadata:
name: ecoemploi
- ConfigMap:
apiVersion: v1
kind: ConfigMap
metadata:
name: postgres-config
namespace: ecoemploi
labels:
app: sources
data:
POSTGRES_DB: postgresdb
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
- Persistent Volume and Persistent Volume Claim:
kind: PersistentVolume
apiVersion: v1
metadata:
name: postgres-pv-volume
namespace: ecoemploi
labels:
type: local
app: sources
spec:
storageClassName: manual
capacity:
storage: 50Gi
accessModes:
- ReadWriteMany
hostPath:
path: "/mnt/data"
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: postgres-pv-claim
namespace: ecoemploi
labels:
app: sources
spec:
storageClassName: manual
accessModes:
- ReadWriteMany
resources:
requests:
storage: 50Gi
- PostGIS sgbd:
The image used, postgis/postgis:14-3.3, is the one I was using on the Dockerfile.
apiVersion: apps/v1
kind: Deployment
metadata:
name: postgres
namespace: ecoemploi
spec:
replicas: 1
selector:
matchLabels:
app: sources
template:
metadata:
labels:
app: sources
spec:
containers:
- name: sources
image: postgis/postgis:14-3.3
imagePullPolicy: "IfNotPresent"
ports:
- containerPort: 5435
envFrom:
- configMapRef:
name: postgres-config
volumeMounts:
- mountPath: /var/lib/postgresql/data
name: ecoemploigis
volumes:
- name: ecoemploigis
persistentVolumeClaim:
claimName: postgres-pv-claim
- Service:
apiVersion: v1
kind: Service
metadata:
name: postgres
namespace: ecoemploi
labels:
app: sources
spec:
type: NodePort
ports:
- port: 5435
selector:
app: sources
Here's what Minikube responds to a kubectl get all --namespace=ecoemploi:
kubectl get all --namespace=ecoemploi
NAME READY STATUS RESTARTS AGE
pod/postgres-687c6f59f6-8fkb4 1/1 Running 2 (10h ago) 15h
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/postgres NodePort 10.98.16.113 <none> 5435:31956/TCP 15h
NAME READY UP-TO-DATE AVAILABLE AGE
deployment.apps/postgres 1/1 1 1 15h
NAME DESIRED CURRENT READY AGE
replicaset.apps/postgres-687c6f59f6 1 1 1 15h
According to the example I followed on the Internet, what I have should allow me to execute a psql command into that running pod, with a:
kubectl exec -it pod/postgres-687c6f59f6-8fkb4 --namespace ecoemploi -- psql -h localhost -U postgres --password -p 5435 ecoemploigis
But psql fails, with this message:
psql: error: connection to server at "localhost" (127.0.0.1), port 5435 failed: Connection refused
Is the server running on that host and accepting TCP/IP connections?
connection to server at "localhost" (::1), port 5435 failed: Cannot assign requested address
Is the server running on that host and accepting TCP/IP connections?
command terminated with exit code 2
Executing a kubectl exec -it pod/postgres-687c6f59f6-8fkb4 --namespace ecoemploi -- bash instead, leads me to the pod's internal, where these folders are present:
bin boot dev docker-entrypoint-initdb.d etc home lib lib64 media mnt opt proc root run sbin srv sys tmp usr var
but /var/log content doesn't explain anything:
-rw-r--r-- 1 root root 8436 Mar 18 16:59 alternatives.log
drwxr-xr-x 1 root root 4096 Mar 18 16:59 apt
-rw-rw---- 1 root utmp 0 Feb 27 00:00 btmp
-rw-r--r-- 1 root root 70686 Mar 18 16:59 dpkg.log
-rw-r--r-- 1 root root 32000 Mar 1 13:53 faillog
-rw-rw-r-- 1 root utmp 292000 Mar 1 13:53 lastlog
drwxrwxr-t 2 root postgres 4096 Feb 9 10:28 postgresql
-rw-rw-r-- 1 root utmp 0 Feb 27 00:00 wtmp
(only dpkg.log and alternatives.log really changed)
and I have:
ls -l /var/log/postgresql/
total 0
Things look like Kubernetes didn't really attempted to run my ecoemploi "sub-system".
It's my first Kubernetes deployment. Am I missing something important?
