K8s - Cannot access service/clusterIP from another node - kubernetes

Question

I am trying to access(from worker node) a pod(on worker node) via a service/ClusterIP using curl http://cluster_ip:port_no but it isn't working.
Here's some info on service
masternode#Master:/localdocker$ kubectl describe svc registry
Name: registry
Namespace: default
Labels: io.kompose.service=registry
Annotations: kompose.cmd: kompose convert -f docker-compose.yaml -o localregistry.yaml
kompose.version: 1.1.0 (36652f6)
Selector: io.kompose.service=registry
Type: ClusterIP
IP: 10.100.126.230
Port: 5000 5000/TCP
TargetPort: 5000/TCP
Endpoints: 192.168.171.74:5000
Session Affinity: None
Events: <none>```
here's some info on pod
masternode#Master:/localdocker$ kubectl describe pod registry-7ccd695dc7-69cx4
Name: registry-7ccd695dc7-69cx4
Namespace: default
Priority: 0
Node: worker/10.0.1.5
Start Time: Sun, 19 Jul 2020 06:09:14 +0000
Labels: io.kompose.service=registry
pod-template-hash=7ccd695dc7
Annotations: cni.projectcalico.org/podIP: 192.168.171.74/32
cni.projectcalico.org/podIPs: 192.168.171.74/32
Status: Running
IP: 192.168.171.74
IPs:
IP: 192.168.171.74
Controlled By: ReplicaSet/registry-7ccd695dc7
Containers:
registry:
Container ID: docker://ca372f12ef7a1a3cb23e7d6c58337f47848f91212f6c75af6bfd04bc48ea2f27
Image: registry:2
Image ID: docker-pullable://registry#sha256:8be26f81ffea54106bae012c6f349df70f4d5e7e2ec01b143c46e2c03b9e551d
Port: 5000/TCP
Host Port: 0/TCP
State: Running
Started: Sun, 19 Jul 2020 06:09:24 +0000
Ready: True
Restart Count: 0
Environment:
REGISTRY_STORAGE_FILESYSTEM_ROOTDIRECTORY: /data
Mounts:
/data from registry-claim0 (rw)
/var/run/secrets/kubernetes.io/serviceaccount from default-token-fhf5k (ro)
Conditions:
Type Status
Initialized True
Ready True
ContainersReady True
PodScheduled True
Volumes:
registry-claim0:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: registry-claim0
ReadOnly: false
default-token-fhf5k:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-fhf5k
Optional: false
QoS Class: BestEffort
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal Scheduled 13m default-scheduler Successfully assigned default/registry-7ccd695dc7-69cx4 to worker
Normal Pulling 13m kubelet, worker Pulling image "registry:2"
Normal Pulled 13m kubelet, worker Successfully pulled image "registry:2"
Normal Created 13m kubelet, worker Created container registry
Normal Started 13m kubelet, worker Started container registry
This is a practice exercise where I was able to do so(in their live environment) without any NodePorts.
Please let me know if any other info is required.

Answer 1

This is an expected behavior because ClusterIP type service is only accessible from within the kubernetes cluster i.e from another pod etc.
If you want to access a pod via a service from outside the kubernetes cluster i.e from the nodes itself then use NodePort type service.
Once you expose it via NodePort service you would be able to access it using curl http://<NODE-IP>:<NODE-PORT>
ClusterIP is created on service network of the cluster and nodes are in different network. By creating a NodePort service a Port is opened in each nodes network to forward traffic to ClusterIP. So in essence NodePort uses ClusterIP internally and is an higher level abstraction built on top of ClusterIP.