#Resource Quotas

Refer to the official Kubernetes documentation: Resource Quotas

#Resource Requests and Resource Limits

This is used to restrict the resources available for a designated namespace. The total resource usage by all Pods in a namespace, except those in a non-terminating state, cannot exceed this value.

• Resource Requests define the minimum amount of resources (e.g., CPU, memory) a container requires, guiding the Kubernetes Scheduler to place the Pod on a Node with sufficient resources.
• Resource Limits set the maximum amount of resources a container can consume, preventing resource exhaustion and ensuring system stability.

#Quotas

#Resource Quotas

Unlimited means that there is no restriction on input and usage, but actual usage cannot exceed the available value of the corresponding resource in the cluster.

Each container’s computing resource limit (limits) specified under the namespace, or each new Pod or PVC added, will consume the quota set here.

Includes the following resource quotas

Field	Description
Resource Requests	Specifies the quota value of the resource requests. CPU Memory
Resource Limits	Specifies the quota value of the resource limits. CPU Memory
Number of Pods	Specifies the quota number of Pods.

Note:

• The resource quota of the namespace comes from the quota allocated to the project by the cluster. For any given resource, the maximum quota that can be set is the unused quota in the project. If the available amount of any resource is 0, you will not be able to create a namespace. Please contact the platform administrator for support.
• GPU resource quota can only be configured when GPU (vGPU or pGPU) resources are deployed in the cluster. When vGPU resources are used, vMemory quota can also be configured. GPU Units: 100 units of virtual cores are equivalent to 1 physical core (1 pGPU = 1 core = 100 vGPU), and pGPU is allocated in whole units. 1 unit of memory is equal to 256 MiB, and 1024 MiB = 1 GiB.
• If no quota is specified for a certain resource, it defaults to Unlimited. This means that the namespace can use the available resources of the corresponding type allocated to the project in the cluster as needed, without any limit.

#Storage Quotas

Quota Type:

• All: The total requested storage capacity of all Persistent Volume Claims (PVCs) in this namespace cannot exceed this value.
• Storage Class: In this namespace, the total requested storage capacity of all Persistent Volume Claims (PVCs) associated with the selected storage class cannot exceed this value.

Note: Please allocate the storage class to the project where the namespace resides in advance.

#Extended Resources Quotas

The quota types for extended resources are obtained from the configuration dictionary (ConfigMap). If the corresponding configuration dictionary does not exist, this category will not be displayed.

Configuration Dictionary Field Descriptions

Field	Description
data.dataType	Specifies the data type of the value.
data.defaultValue	Sets the default value. If this field does not exist or the value is empty, it indicates that there is no default value.
data.descriptionEn	English prompt message, which can be viewed by clicking the icon on the right side of the value.
data.descriptionZh	Chinese prompt message, which can be viewed by clicking the icon on the right side of the value.
data.excludeResources	Mutually exclusive resources. This field is used to add resources that cannot be configured simultaneously, separated by ,.
data.group	Group
data.groupI18n	The English and Chinese names of the group, which can be viewed and selected in the drop-down options corresponding to the quota types. Applicable when the same resource has one or more values.
data.key	Specifies the value of the key. A configuration dictionary can only describe one key.
data.labelEn/data.labelZh	The English/Chinese name of the resource, which can be viewed and selected in the drop-down options corresponding to the quota types. This field serves the same function as the data.groupI18n field but is only applicable when the same resource has a single value, ensuring compatibility with the old version of the configuration dictionary (ConfigMap).
data.limits	Indicates whether to configure limits for the resources. Valid values include: disabled indicates limits cannot be configured for the resource, required indicates it must be input, and optional indicates it is optional input.
data.requests	Indicates whether to configure requests for the resources. Valid values include: disabled indicates requests cannot be configured for the resource, required indicates it must be input, optional indicates it is optional input, and fromLimits indicates it will use the same configuration as limits.
data.relatedResources	Associated resources. This field is reserved and currently cannot be used.
data.resourceUnit	Unit of the resource, must be input in English letters, and currently does not support input in Chinese.
data.runtimeClassName	Required runtime class, defaults to empty with optional value nvidia.
metadata.labels	features.cpaas.io/type: CustomResourceLimitation, this label is mandatory and cannot be changed. features.cpaas.io/group: <*groupName*>, this label is mandatory and should modify groupName to the corresponding group. features.cpaas.io/enabled: "true", this label is mandatory and indicates whether it is enabled, default is true.
metadata.name	The format is cf-crl-<*groupName*>-<*name*>, where cf-crl is a fixed field and cannot be changed. groupName is the name of the corresponding resource group, e.g., gpu-manager, galaxy, etc. name is the resource name: Resource name can be standard resource type names, e.g., cpu, memory, pods, etc. The standard resource names must comply with Kubernetes' qualified name rules and must exist within the defined standard resource types in Kubernetes. Resource names can also be special resource types starting with specific prefixes, such as: hugepages- or requests.hugepages-.
metadata.namespace	Must be specified as the kube-public namespace and cannot be changed.

Configuration Dictionary Example

This document only provides YAML configuration examples for physical GPU, GPU Manager core count and memory, MPS core count and memory. For using other resources, please refer to the above field descriptions for customization.

• Physical GPU Configuration Example

  apiVersion: v1
  data:
    dataType: integer
    defaultValue: "1"
    descriptionEn: GPU quantity
    descriptionZh: GPU 数量
    excludeResources: nvidia.com/mps-core,nvidia.com/mps-memory,tencent.com/vcuda-core,tencent.com/vcuda-memory
    group: nvidia-device-plugin
    groupI18n: '{"zh": "pGPU", "en": "pGPU"}'
    key: nvidia.com/gpu
    labelEn: GPU core
    labelZh: 物理 GPU
    limits: optional
    relatedResources: ""
    requests: disabled
    resourceUnit: count
    runtimeClassName: nvidia
  kind: ConfigMap
  metadata:
    labels:
      features.cpaas.io/enabled: "true"
      features.cpaas.io/group: nvidia-device-plugin
      features.cpaas.io/type: CustomResourceLimitation
    name: cf-crl-nvidia-device-plugin
    namespace: kube-public
  

• GPU Manager Configuration Example

  • GPU Manager Core Count:

    apiVersion: v1
    data:
      dataType: integer
      defaultValue: "20"
      descriptionEn: "GPU vcore count, 100 vcore equals 1 GPU core"
      descriptionZh: 虚拟 GPU 核心数，100个虚拟核心等于 1 个 GPU 物理核心
      excludeResources: nvidia.com/gpu,nvidia.com/mps-core,nvidia.com/mps-memory
      group: gpu-manager
      groupI18n: '{"zh": "GPU Manager", "en": "GPU Manager"}'
      key: tencent.com/vcuda-core
      labelEn: GPU vcore
      labelZh: 虚拟 GPU
      limits: optional
      relatedResources: tencent.com/vcuda-memory
      requests: disabled
      resourceUnit: vcore
      runtimeClassName: ""
    kind: ConfigMap
    metadata:
      labels:
        device-recource.cpaas.io/config: "true"
        features.cpaas.io/enabled: "true"
        features.cpaas.io/group: gpu-manager
        features.cpaas.io/type: CustomResourceLimitation
      name: demo-gpu-manager-core
      namespace: kube-public
    

  • GPU Manager Memory:

    apiVersion: v1
    data:
      dataType: integer
      defaultValue: "20"
      descriptionEn: GPU vmemory, 256 Mi physical memory per Unit
      descriptionZh: 虚拟 GPU 显存，每单位等于 256Mi 显存
      excludeResources: nvidia.com/gpu,nvidia.com/mps-core,nvidia.com/mps-memory
      group: gpu-manager
      groupI18n: '{"zh": "GPU Manager", "en": "GPU Manager"}'
      key: tencent.com/vcuda-memory
      labelEn: GPU vmemory
      labelZh: 显存
      limits: optional
      relatedResources: tencent.com/vcuda-core
      requests: disabled
      resourceUnit: unit
      runtimeClassName: ""
    kind: ConfigMap
    metadata:
      labels:
        device-recource.cpaas.io/config: "true"
        features.cpaas.io/enabled: "true"
        features.cpaas.io/group: gpu-manager
        features.cpaas.io/type: CustomResourceLimitation
      name: demo-gpu-manager-memory
      namespace: kube-public
    

• MPS Configuration Example

  • MPS Core Count:

    apiVersion: v1
    data:
      dataType: integer
      defaultValue: "20"
      descriptionEn: "MPS GPU core count, 100 MPS core equals 1 GPU core"
      descriptionZh: MPS GPU 核心数，100个 MPS 核心等于 1 个 GPU 物理核心
      excludeResources: nvidia.com/gpu,tencent.com/vcuda-core,tencent.com/vcuda-memory
      group: nvidia-mps-device-plugin
      groupI18n: '{"zh": "MPS", "en": "MPS"}'
      key: nvidia.com/mps-core
      labelEn: MPS GPU core
      labelZh: MPS GPU
      limits: optional
      relatedResources: nvidia.com/mps-memory
      requests: disabled
      resourceUnit: vcore
      runtimeClassName: nvidia
    kind: ConfigMap
    metadata:
      labels:
        features.cpaas.io/enabled: "true"
        features.cpaas.io/group: nvidia-mps-device-plugin
        features.cpaas.io/type: CustomResourceLimitation
      name: cf-crl-nvidia-mps-device-plugin-core
      namespace: kube-public
    

  • MPS Memory:

    apiVersion: v1
    data:
      dataType: integer
      defaultValue: "1"
      descriptionEn: MPS GPU memory, 1 Gi physical memory per Unit
      descriptionZh: MPS GPU 显存，每单位等于 1Gi 显存
      excludeResources: nvidia.com/gpu,tencent.com/vcuda-core,tencent.com/vcuda-memory
      group: nvidia-mps-device-plugin
      groupI18n: '{"zh": "MPS", "en": "MPS"}'
      key: nvidia.com/mps-memory
      labelEn: MPS GPU memory
      labelZh: MPS 显存
      limits: optional
      relatedResources: nvidia.com/mps-core
      requests: disabled
      resourceUnit: unit
      runtimeClassName: nvidia
    kind: ConfigMap
    metadata:
      labels:
        features.cpaas.io/enabled: "true"
        features.cpaas.io/group: nvidia-mps-device-plugin
        features.cpaas.io/type: CustomResourceLimitation
      name: cf-crl-nvidia-mps-device-plugin-memory
      namespace: kube-public
    

#Other Quotas

The format for custom quota names must comply with the following specifications:

• If the custom quota name does not contain a slash (/): It must start and end with a letter or number, and can contain letters, numbers, hyphens (-), underscores (_), or periods (.), forming a qualified name with a maximum length of 63 characters.
• If the custom quota name contains a slash (/): The name is divided into two parts: prefix and name, in the form of: prefix/name. The prefix must be a valid DNS subdomain, while the name must comply with the rules for a qualified name.
• DNS Subdomain:
  • Label: Must start and end with lowercase letters or numbers, may contain hyphens (-), but cannot be exclusively composed of hyphens, with a maximum length of 63 characters.
  • Subdomain: Extends the rules of the label, allowing multiple labels to be connected by periods (.) to form a subdomain, with a maximum length of 253 characters.