perf: request limit

Author: c121914yu

document/content/docs/upgrading/4-14/4143.mdx

@@ -9,6 +9,7 @@ description: 'FastGPT V4.14.2 更新说明'
 1. 全局变量支持文件上传
 2. 知识库文件迁移至 S3(全部使用文件的地方均已迁移)
 3. 系统插件，插件市场中会提示是否有新版本，并提供更新按键。
+4. 工作流运行 QPM 限制。
 
 ## ⚙️ 优化
 

document/data/doc-last-modified.json

@@ -117,7 +117,7 @@
   "document/content/docs/upgrading/4-14/4140.mdx": "2025-11-06T15:43:00+08:00",
   "document/content/docs/upgrading/4-14/4141.mdx": "2025-11-19T10:15:27+08:00",
   "document/content/docs/upgrading/4-14/4142.mdx": "2025-11-18T19:27:14+08:00",
-  "document/content/docs/upgrading/4-14/4143.mdx": "2025-11-21T16:42:49+08:00",
+  "document/content/docs/upgrading/4-14/4143.mdx": "2025-11-21T18:05:13+08:00",
   "document/content/docs/upgrading/4-8/40.mdx": "2025-08-02T19:38:37+08:00",
   "document/content/docs/upgrading/4-8/41.mdx": "2025-08-02T19:38:37+08:00",
   "document/content/docs/upgrading/4-8/42.mdx": "2025-08-02T19:38:37+08:00",

packages/service/common/api/frequencyLimit.ts

@@ -1,24 +1,28 @@
+/* 基于 Team 的限流 */
 import { getGlobalRedisConnection } from '../../common/redis';
 import { jsonRes } from '../../common/response';
 import type { NextApiResponse } from 'next';
 
+export enum LimitTypeEnum {
+  chat = 'chat'
+}
+const limitMap = {
+  [LimitTypeEnum.chat]: {
+    seconds: 60,
+    limit: Number(process.env.CHAT_MAX_QPM || 5000)
+  }
+};
+
 type FrequencyLimitOption = {
   teamId: string;
-  seconds: number;
-  limit: number;
-  keyPrefix: string;
+  type: LimitTypeEnum;
   res: NextApiResponse;
 };
 
-export const teamFrequencyLimit = async ({
-  teamId,
-  seconds,
-  limit,
-  keyPrefix,
-  res
-}: FrequencyLimitOption) => {
+export const teamFrequencyLimit = async ({ teamId, type, res }: FrequencyLimitOption) => {
+  const { seconds, limit } = limitMap[type];
   const redis = getGlobalRedisConnection();
-  const key = `${keyPrefix}:${teamId}`;
+  const key = `frequency:${type}:${teamId}`;
 
   const result = await redis
     .multi()

packages/service/type/env.d.ts

@@ -32,6 +32,7 @@ declare global {
       PASSWORD_LOGIN_LOCK_SECONDS?: string;
       PASSWORD_EXPIRED_MONTH?: string;
       MAX_LOGIN_SESSION?: string;
+      CHAT_MAX_QPM?: string;
 
       // 安全配置
       // 密码登录锁定时间

projects/app/.env.template

@@ -96,6 +96,8 @@ PASSWORD_LOGIN_LOCK_SECONDS=
 PASSWORD_EXPIRED_MONTH=
 # 最大登录客户端数量，默认为 10
 MAX_LOGIN_SESSION=
+# 工作流QPM
+CHAT_MAX_QPM=5000
 
 # 特殊配置
 # 自定义跨域，不配置时，默认都允许跨域（逗号分割）

projects/app/src/pages/api/core/chat/chatTest.ts

@@ -49,7 +49,7 @@ import {
 import { saveChat, updateInteractiveChat } from '@fastgpt/service/core/chat/saveChat';
 import { getLocale } from '@fastgpt/service/common/middle/i18n';
 import { formatTime2YMDHM } from '@fastgpt/global/common/string/time';
-import { clone } from 'lodash';
+import { LimitTypeEnum, teamFrequencyLimit } from '@fastgpt/service/common/api/frequencyLimit';
 
 export type Props = {
   messages: ChatCompletionMessageParam[];
@@ -93,6 +93,16 @@ async function handler(req: NextApiRequest, res: NextApiResponse) {
       per: ReadPermissionVal
     });
 
+    if (
+      !(await teamFrequencyLimit({
+        teamId,
+        type: LimitTypeEnum.chat,
+        res
+      }))
+    ) {
+      return;
+    }
+
     const isPlugin = app.type === AppTypeEnum.workflowTool;
     const isTool = app.type === AppTypeEnum.tool;
 

projects/app/src/pages/api/v1/chat/completions.ts

@@ -60,8 +60,7 @@ import { getWorkflowToolInputsFromStoreNodes } from '@fastgpt/global/core/app/to
 import { UserError } from '@fastgpt/global/common/error/utils';
 import { getLocale } from '@fastgpt/service/common/middle/i18n';
 import { formatTime2YMDHM } from '@fastgpt/global/common/string/time';
-import { teamFrequencyLimit } from '@fastgpt/service/common/api/frequencyLimit';
-import { clone } from 'lodash';
+import { LimitTypeEnum, teamFrequencyLimit } from '@fastgpt/service/common/api/frequencyLimit';
 
 type FastGptWebChatProps = {
   chatId?: string; // undefined: get histories from messages, '': new chat, 'xxxxx': get histories from db
@@ -190,14 +189,13 @@ async function handler(req: NextApiRequest, res: NextApiResponse) {
     if (
       !(await teamFrequencyLimit({
         teamId,
-        keyPrefix: 'chat:completions',
-        seconds: 60,
-        limit: 5000,
+        type: LimitTypeEnum.chat,
         res
       }))
     ) {
-      return {};
+      return;
     }
+
     retainDatasetCite = retainDatasetCite && !!responseDetail;
     const isPlugin = app.type === AppTypeEnum.workflowTool;
 

projects/app/src/pages/api/v2/chat/completions.ts

@@ -61,7 +61,7 @@ import { getWorkflowToolInputsFromStoreNodes } from '@fastgpt/global/core/app/to
 import { UserError } from '@fastgpt/global/common/error/utils';
 import { getLocale } from '@fastgpt/service/common/middle/i18n';
 import { formatTime2YMDHM } from '@fastgpt/global/common/string/time';
-import { clone } from 'lodash';
+import { LimitTypeEnum, teamFrequencyLimit } from '@fastgpt/service/common/api/frequencyLimit';
 
 type FastGptWebChatProps = {
   chatId?: string; // undefined: get histories from messages, '': new chat, 'xxxxx': get histories from db
@@ -186,6 +186,17 @@ async function handler(req: NextApiRequest, res: NextApiResponse) {
         chatId
       });
     })();
+
+    if (
+      !(await teamFrequencyLimit({
+        teamId,
+        type: LimitTypeEnum.chat,
+        res
+      }))
+    ) {
+      return;
+    }
+
     retainDatasetCite = retainDatasetCite && !!responseDetail;
     const isPlugin = app.type === AppTypeEnum.workflowTool;