scheduler: consider pod requests when gpu&RDMA joint allocate

[ZiMengSheng]

Ⅰ. Describe what this PR does

a node with two numaNode, two pcie per numaNode, two gpu and two rdma per pcie, user may want two gpu and two rdma by following protocol

apiVersion: v1
kind: Pod
metadata:
  name: hello-gpu
  namespace: default
  annotations:
    # 表示分走 deviceTypes[0] 之后会尝试在同 PCIE 继续分配 deviceTypes[1]、deviceTypes[2]、...
    scheduling.koordinator.sh/device-joint-allocate:|
      {
        "deviceTypes": [
            "gpu",
            "rdma"
        ]
      }
spec:
  containers:
  - resources:
      limits:
        cpu: "16"
        memory: 32Gi
        nvidia.com/gpu: "2"
        koordinator.sh/rdma: "200"
      requests:
        cpu: "16"
        memory: 32Gi
        nvidia.com/gpu: "2"
        koordinator.sh/rdma: "200"

Ⅱ. Does this pull request fix one issue?

Ⅲ. Describe how to verify it

Ⅳ. Special notes for reviews

V. Checklist

• I have written necessary docs and comments
• I have added necessary unit tests and integration tests
• All checks passed in make test

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 66.29%. Comparing base (63f9c72) to head (73d8080).
Report is 2 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2233   +/-   ##
=======================================
  Coverage   66.29%   66.29%           
=======================================
  Files         453      453           
  Lines       53175    53175           
=======================================
+ Hits        35250    35255    +5     
+ Misses      15389    15382    -7     
- Partials     2536     2538    +2     

Flag	Coverage Δ
unittests	66.29% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[songtao98]

/lgtm

[jasonliu747]

/lgtm