> For the complete documentation index, see [llms.txt](https://tccli-agent.gitbook.io/tccli/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://tccli-agent.gitbook.io/tccli/v2/tke-rong-qi-fu-wu/quotas.md).

# 配额和限制

> 资源配额与 API 限频。**TKE 资源配额默认值**以官方 [购买集群配额限制](https://cloud.tencent.com/document/product/457/9087) 为准（可提工单调高）；API 限频以 TKE API 概览为准。超限信号列给出触发限制时看到的错误码。下列数值为官方默认配额，调额后以官网最新表为准。

## 资源配额

| 限制项      |        默认值       | 作用域    |               可调整              | 超限信号                                 |
| -------- | :--------------: | ------ | :----------------------------: | ------------------------------------ |
| 单地域集群数   |      **20**      | 每地域每账号 |             是（提工单）             | `InternalError.QuotaMaxClsLimit`     |
| 单集群节点数   |     **5000**     | 每集群    |             是（提工单）             | `InternalError.QuotaMaxNodLimit`     |
| 节点池数     |        20        | 每集群    |             是（提工单）             | `LimitExceeded`                      |
| 每节点池节点数  | 受 ASG MaxSize 限制 | 每节点池   |            是（改 ASG）            | 扩容时 ASG 达上限                          |
| 重启节点上限   |        100       | 每次请求   |                否               | `InvalidParameterValue`              |
| 集群规格（等级） |    L5 \~ L5000   | 每集群    | 是（`ModifyClusterLevel` / 自动升配） | `InvalidParameterValue.ClusterLevel` |

> **容器网段 CIDR**：创建时自定义；决定节点上限 / Service 上限 / 每节点 Pod 上限；**暂不支持变更**（变更须重建集群，或走受限路径如 GR 的 `AddClusterCIDR`）。

### 集群规格与 K8s 资源配额（按规格） <a href="#ji-qun-gui-ge-yu-k8s-zi-yuan-peiean-gui-ge" id="ji-qun-gui-ge-yu-k8s-zi-yuan-peiean-gui-ge"></a>

> 下列配额自 2022-04-30 起生效，无法移除；不足时升高集群规格。超级节点 Pod 与非超级节点 Pod **分列**限制。

|  集群规格 | Pod 上限（非超级节点） | Pod 上限（超级节点） | ConfigMap 上限 | CRD / 其他 K8s 资源上限 |
| :---: | :-----------: | :----------: | :----------: | :---------------: |
|   L5  |      600      |      400     |      256     |        1250       |
|  L20  |      1500     |      800     |      512     |        2500       |
|  L50  |      3000     |     1200     |     1024     |        5000       |
|  L100 |      6000     |     1600     |     2048     |       10000       |
|  L200 |     15000     |     2000     |     4096     |       20000       |
|  L500 |     30000     |     4000     |     6144     |       50000       |
| L1000 |     90000     |     8000     |     8192     |       100000      |
| L3000 |     150000    |     24000    |     10240    |       150000      |
| L5000 |     200000    |     40000    |     20480    |       200000      |

> 集群规格决定管理费与可管理规模；选型见 [创建集群](/tccli/v2/tke-rong-qi-fu-wu/index-1/create.md) 的等级与容量规划说明。计费项拆分见官方计费概述（默认配额与计费数值以官网最新表为准）。

## API 限频

> 限频值见腾讯云 TKE API 概览页。多数接口 20 次/秒，写操作与高频查询有差异。

| 接口类别                                                                      | 限频 (次/秒) | 超限信号                   |
| ------------------------------------------------------------------------- | :------: | ---------------------- |
| `DescribeClusters` / `DescribeClusterStatus` / `DescribeClusterNodePools` |    20    | `RequestLimitExceeded` |
| `DescribeClusterEndpoints` / `DescribeClusterKubeconfig`                  |    20    | `RequestLimitExceeded` |
| `CreateCluster` / `CreateClusterNodePool`                                 |    20    | `RequestLimitExceeded` |
| `DeleteCluster`                                                           |    50    | `RequestLimitExceeded` |
| `DescribeClusterSecurity`                                                 |    100   | `RequestLimitExceeded` |
| 边缘集群查询（`DescribeEdgeClusterInstances` 等）                                  |    10    | `RequestLimitExceeded` |

> 通用默认限频 10 次/秒，TKE 多数接口放宽到 20 次/秒。批量操作时串行调用或加间隔，避免触发 `RequestLimitExceeded`。

## 查询配额用量

```bash
# 集群存量 (核对是否接近单地域默认 20 上限)
tccli tke DescribeClusters --region <REGION> --output text --filter "TotalCount"
# expected: 数字 ≤ 20（默认配额，已提额账号可更高）

# 节点池存量 (核对是否接近 20 上限)
tccli tke DescribeClusterNodePools --region <REGION> --ClusterId "<CLUSTER_ID>" --output text --filter "TotalCount"
# expected: 数字 ≤ 20

# 集群节点数 (核对 ClusterRunningNodeNum)
tccli tke DescribeClusterStatus --region <REGION> --filter "ClusterStatusSet[?ClusterId=='<CLUSTER_ID>'] | [0].{state:ClusterState,running:ClusterRunningNodeNum,init:ClusterInitNodeNum}"
# expected: running 节点数 ≤ 5000
```

## 资源用量查询

> 集群资源用量（CRD/节点）查询，核对是否接近配额上限。

```bash
# 集群资源用量 (CRD/Pod/RS/ConfigMap 等存量, 核对是否接近上限)
tccli tke DescribeResourceUsage --ClusterId "<CLUSTER_ID>" --region <REGION>
# expected: exit 0, 返回 5 个用量字段: CRDUsage/PodUsage/RSUsage/ConfigMapUsage/OtherUsage
```

```json
{
    "CRDUsage": {"Name": "CRD", "Usage": 10, "Details": [{"Name": "example-crd.example.com", "Usage": 2}]},
    "PodUsage": 5,
    "RSUsage": 3,
    "ConfigMapUsage": 2,
    "OtherUsage": {"Name": "Other", "Usage": 1, "Details": [{"Name": "services", "Usage": 1}]}
}
```

```bash
# 节点资源 (含预留实例, 需 NodeName)
tccli tke DescribePostNodeResources --ClusterId "<CLUSTER_ID>" --NodeName "<NODE_NAME>" --region <REGION>
# expected: exit 0, ReservedInstanceSet[]
```

> `DescribeResourceUsage` 返回 5 类资源存量：`CRDUsage`/`OtherUsage` 为对象（`Name`/`Usage`/`Details[]`），`PodUsage`/`RSUsage`/`ConfigMapUsage` 为整数。核对是否接近 [集群规格配额表](#集群规格与-k8s-资源配额按规格)。`DescribePostNodeResources` 查节点资源含预留实例绑定。

## Pod 成本核算与预留实例抵扣

> Pod 计费、按规格估算可创建数、预留实例抵扣率与利用率——成本核算任务，区别于资源配额与用量查询。

```bash
# 按规格查询 Pod (CPU/Mem/GPU 规格 + 可用区, 创建前估算)
tccli tke DescribePodsBySpec --Cpu 2 --Memory 4 --Zone "<ZONE>" --ClusterId "<CLUSTER_ID>" --region <REGION>
# expected: exit 0, 可创建 Pod 数

# Pod 计费信息 (按 Namespace+Name 或 Uids)
tccli tke DescribePodChargeInfo --ClusterId "<CLUSTER_ID>" --Namespace "<NS>" --Name "<POD_NAME>" --region <REGION>
# expected: exit 0, Pod 计费详情

# Pod 抵扣率 (按可用区, 预留实例抵扣情况)
tccli tke DescribePodDeductionRate --Zone "<ZONE>" --ClusterId "<CLUSTER_ID>" --region <REGION>
# expected: exit 0, 抵扣率

# 预留实例利用率明细 (查询, 不绑集群)
tccli tke DescribeRIUtilizationDetail --Limit 20 --region <REGION>
# expected: exit 0, 预留实例利用明细
```

> `DescribePodsBySpec` 按规格估算可创建 Pod 数。`DescribePodChargeInfo`/`DescribePodDeductionRate` 是 Pod 计费与预留实例抵扣。`DescribeRIUtilizationDetail` 查询预留实例利用率。
>
> **预留实例范围**：预留实例的**写操作**（购买/退还/续费/改范围）见 [本指南不覆盖哪些操作](https://tccli-agent.gitbook.io/tccli/v2/tke-rong-qi-fu-wu/pages/efhPLsIy1TdooHGCgkj6#本指南不覆盖哪些操作)；预留实例的**查询用量**（`DescribeRIUtilizationDetail`）与 Pod 成本核算同属成本核对任务，由本页覆盖。

## 相关文档

* [错误码](/tccli/v2/tke-rong-qi-fu-wu/error-codes.md) — `LimitExceeded` / `QuotaMaxClsLimit` 的诊断
* [创建集群](/tccli/v2/tke-rong-qi-fu-wu/index-1/create.md) — 集群数配额检查
* [创建节点池](/tccli/v2/tke-rong-qi-fu-wu/index-2/nodepool-create.md) — 节点池配额检查


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://tccli-agent.gitbook.io/tccli/v2/tke-rong-qi-fu-wu/quotas.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.