[SOLVED] node is assigning IP to pods that is outside of specified CIDRs

$ kubectl describe nodes lke17110-20819-607cabf9bf93

System Info:
  Machine ID:                                   a0c04b65ec6349daa7d0c11997713e5e
  System UUID:                                  a0c04b65ec6349daa7d0c11997713e5e
  Boot ID:                                      053968c1-3610-4172-b62e-320976fe709c
  Kernel Version:                               5.10.0-5-cloud-amd64
  OS Image:                                     Debian GNU/Linux 9 (stretch)
  Operating System:                             linux
  Architecture:                                 amd64
  Container Runtime Version:                    docker://19.3.15
  Kubelet Version:                              v1.20.5
  Kube-Proxy Version:                           v1.20.5
PodCIDR:                                        10.2.5.0/24
PodCIDRs:                                       10.2.5.0/24
ProviderID:                                     linode://26132501
Non-terminated Pods:                            (7 in total)
  Namespace                                     Name                                                    CPU Requests  CPU Limits  Memory Requests  Memory Limits  Age
  ---------                                     ----                                                    ------------  ----------  ---------------  -------------  ---
  default                                       ngx-ingress-ingress-nginx-controller-6c9496447-f4dmd    100m (10%)    0 (0%)      90Mi (4%)        0 (0%)         121d
  redacted-project-12-review-k8s-dev-pe-cl5zoc  postgres-1                                              0 (0%)        0 (0%)      230Mi (12%)      230Mi (12%)    42m
  kube-system                                   calico-node-n7nck                                       250m (25%)    0 (0%)      0 (0%)           0 (0%)         27d
  kube-system                                   csi-linode-node-k7n9v                                   0 (0%)        0 (0%)      0 (0%)           0 (0%)         12d
  kube-system                                   kube-proxy-9n4s2                                        0 (0%)        0 (0%)      0 (0%)           0 (0%)         27d
  kube-system                                   metrics-server-d895c4b8b-bglm6                          0 (0%)        0 (0%)      0 (0%)           0 (0%)         121d

i.e. Pod IP range should be within 10.2.5.0/24 for this node. BUT:

$ kubectl get pods -n redacted-project-12-review-k8s-dev-pe-cl5zoc postgres-1 -owide

NAME         READY   STATUS             RESTARTS   AGE   IP             NODE                          NOMINATED NODE   READINESS GATES
postgres-1   0/1     CrashLoopBackOff   13         45m   10.2.173.196   lke17110-20819-607cabf9bf93   <none>           <none>

The actual IP got assigned to postgres-1 is incorrect ( should be 10.2.5.x not 10.2.173.x ). Is this a bug? Since I expect the IP should follow the defined CIDR so I've only allowed such IP to access. Now it is crashing for this specific reason.