minio tenant from operator shows insufficient cpu, but cpu is available

[manju-rn]

Expected Behavior

Minio tenant should get created when doing from operator - via UI

Current Behavior

The stateful set pod of minio fails with following error

manju@k8s:~$ k -n minio1-ns get pods
NAME                                READY   STATUS    RESTARTS   AGE
minio2-pool-0-0                     0/2     Pending   0          4m9s

manju@k8s:~$ k -n minio1-ns describe pod/minio2-pool-0-0
...
Events:
  Type     Reason            Age   From               Message
  ----     ------            ----  ----               -------
  Warning  FailedScheduling  75s   default-scheduler  0/1 nodes are available: 1 Insufficient cpu. preemption: 0/1 nodes are available: 1 No preemption victims found for incoming pod..

but checking the node details (single node cluster), it has sufficient cpu

manju@k8s:~$ k top nodes
NAME   CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%
k8s    666m         16%    4292Mi          17%

Your Environment

• microk8s single node cluster.
• Minio operator is running in the cluster with 1 tenant already running successfully

[manju-rn]

Realized that Allocated CPU requests was already at 97% hence the new Tenant was not scheduling.
However, is there a way to provide "lesser" value during tenant creation - like 100m? Currently the UI only allow whole number as a CPU, but it will be far more flexible to provide discreet values as kubernetes does allow

Allocated resources:
  (Total limits may be over 100 percent, i.e., overcommitted.)
  Resource           Requests      Limits
  --------           --------      ------
  cpu                3900m (97%)   1800m (45%)

I have currently solved this by going to the edit option in UI which shows the YAML file and changed the cpu value from whole number like 1 or 2 (representing absolute CPU cores - 1 CPU - 1000m) to a lessor value like 250m (25% of 1 CPU). This provides more flexibility in assigning computation resources. As my current testing needs only involves around 20m to 100m so wasting 1 full CPU (1000m) bounded by this request is not required and affects my k8s node ability to schedule more pods

 resources:
	requests:
	cpu: 250m
	memory: 2G