LangGraph 1.0 部署大全！三种模式全面覆盖，独立服务器到企业K8s一键搞定！
langgraph 1.0正式版本后，有三种部署的方式， 先说结论：
如果只是部署自己的langgraph服务 使用独立服务器(standalone server)就可以
如果你需要本地的langsmith用来监控，不需要部署管理，使用langsmith方式
如果你还要langsmith的部署管理，使用langsmith with deployment
下面我们分别开始
一 独立服务器(standalone server)
独立服务器：直接部署 LangGraph 服务器，无需控制平面 UI。非常适合将一个或几个代理作为独立服务运行的轻量级设置，并完全控制扩展和集成
1.1 建立LangGraph 应用程序
uv venv --python 3.12
.venvScriptsctivate
uv pip install -U "langgraph-cli[inmem]"12.安装的时候可以直接 langgraph new ./app --template new-langgraph-project-python如果有选择模板的需求， 可以使用 langgraph new 来到下面的交互界面，可以选择模板 以及 语言
📂 Please specify the path to create the application [.]: ./app
🌟 Please select a template:
1. New LangGraph Project - A simple, minimal chatbot with memory.
2. ReAct Agent - A simple agent that can be flexibly extended to many tools.
3. Memory Agent - A ReAct-style agent with an additional tool to store memories for use across conversational threads.
4. Retrieval Agent - An agent that includes a retrieval-based question-answering system.
5. Data-enrichment Agent - An agent that performs web searches and organizes its findings into a structured format.
Enter the number of your template choice (default is 1): 1
You selected: New LangGraph Project - A simple, minimal chatbot with memory.
Choose language (1 for Python 🐍, 2 for JS/TS 🌐): 1
📥 Attempting to download repository as a ZIP archive...
URL: https://github.com/langchain-ai/new-langgraph-project/archive/refs/heads/main.zip
cd app
uv pip install -e .建立一个 .env 里面放上 LANGSMITH_API_KEY=lsv2...
1.3 启动 langgraph server
langgraph dev
1.4 让我们测试下
uv pip install langgraph-sdk
from langgraph_sdk import get_client
import asyncio
client = get_client(url="http://localhost:2024")
async def main():
async for chunk in client.runs.stream(
None, # Threadless run
"agent", # Name of assistant. Defined in langgraph.json.
input={
"messages": [{
"role": "human",
"content": "What is LangGraph?",
}],
},
):
print(f"Receiving new event of type: {chunk.event}...")
print(chunk.data)
print("
")
asyncio.run(main())
1.5 打包成docker image
langgraph build -t langgraph
结束后 docker images 就会看到构建好的镜像
1.6 docker-compose部署
新建一个 docker-compose.yml
volumes:
langgraph-data:
driver: local
services:
langgraph-redis:
image: redis:6
healthcheck:
test: redis-cli ping
interval: 5s
timeout: 1s
retries: 5
langgraph-postgres:
image: postgres:16
ports:
- "5432:5432"
environment:
POSTGRES_DB: postgres
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
volumes:
- langgraph-data:/var/lib/postgresql/data
healthcheck:
test: pg_isready -U postgres
start_period: 10s
timeout: 1s
retries: 5
interval: 5s
langgraph-api:
image: ${IMAGE_NAME}
ports:
- "8123:8000"
depends_on:
langgraph-redis:
condition: service_healthy
langgraph-postgres:
condition: service_healthy
env_file:
- .env
environment:
REDIS_URI: redis://langgraph-redis:6379
LANGSMITH_API_KEY: ${LANGSMITH_API_KEY}
DATABASE_URI: postgres://postgres:postgres@langgraph-postgres:5432/postgres?sslmode=disable
docker compose up 启动
验证下是否启动成功
curl --request GET --url 127.0.0.1:8123/ok
{“ok”:true}
同时langsmith[https://smith.langchain.com/]平台上也可以看到，项目和问答的记录
这份完整版的大模型 AI 学习和面试资料已经上传CSDN，朋友们如果需要可以微信扫描下方CSDN官方认证二维码免费领取【保证100%免费】
二 langsmith 的方式
LangSmith：部署 LangSmith 应用程序的实例，该实例包括 UI 和 API 中的可观测性、跟踪和评估。最适合希望在不部署代理的情况下进行自托管监视和评估的团队
| 服务 | 描述 |
|---|---|
| LangSmith 前端 | 前端使用 Nginx 来提供 LangSmith UI 并将 API 请求路由到其他服务器。这是应用程序的入口点，也是必须向用户公开的唯一组件。 |
| LangSmith 后端 | 后端是 CRUD API 请求的主要入口点，处理应用程序的大部分业务逻辑。这包括处理来自前端和 SDK 的请求、准备用于摄取的跟踪以及支持中心 API。 |
| LangSmith 队列 | 队列处理传入的跟踪和反馈，以确保它们被异步摄取并持久化到跟踪和反馈数据存储中，处理数据完整性检查并确保成功插入到数据存储中，处理数据库错误或暂时无法连接到数据库等情况下的重试。 |
| LangSmith 平台后端 | 平台后端是另一项关键服务，主要处理身份验证、运行摄取和其他大容量任务。 |
| LangSmith 游乐场 | Playground 是一项服务，用于处理向各种 LLM API 转发请求以支持 LangSmith Playground 功能。这也可用于连接到您自己的自定义模型服务器。 |
| LangSmith ACE（任意代码执行）后端 | ACE 后端是一种在安全环境中处理执行任意代码的服务。这用于支持在 LangSmith 中运行自定义代码。 |
仅用于开发/测试。不要将 Docker Compose 用于生产环境。对于生产部署，请使用
Kubernetes。 我这里演示，为了方便使用docker
2.1 生成一个 Api Key Salt
openssl rand -base64 32
例如 U7MfWaDYNzzgMmvJXKn4Pl41mR5jJVHWrhq6c+6jngE=
docker-compose.yml
services:
langchain-playground:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-playground:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
ports:
- 3001:3001
environment:
- PORT=3001
- LANGCHAIN_ENV=local_docker
- LOG_LEVEL=${LOG_LEVEL:-info}
- GO_ENDPOINT=http://langchain-platform-backend:1986
- SMITH_BACKEND_ENDPOINT=${SMITH_BACKEND_ENDPOINT:-http://langchain-backend:1984}
langchain-frontend:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-frontend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
environment:
- VITE_BACKEND_AUTH_TYPE=${AUTH_TYPE:-none}
- VITE_BASIC_AUTH_ENABLED=${BASIC_AUTH_ENABLED:-false}
- VITE_OAUTH_CLIENT_ID=${OAUTH_CLIENT_ID}
- VITE_OAUTH_ISSUER_URL=${OAUTH_ISSUER_URL}
ports:
- 1981:1980
depends_on:
- langchain-backend
- langchain-playground
langchain-ace-backend:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-ace-backend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
ports:
- 1987:1987
environment:
- PORT=1987
command:
- "deno"
- "run"
- "--unstable-worker-options"
- "--allow-env"
- "--allow-net=0.0.0.0:1987"
- "--node-modules-dir"
- "-R"
- "src/main.ts"
- "-R"
- "src/python_worker.ts"
langchain-backend:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-backend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
environment:
- PORT=1984
- LANGCHAIN_ENV=local_docker
- LANGSMITH_URL=${LANGSMITH_URL:-http://langchain-frontend:1980}
- GO_ENDPOINT=http://langchain-platform-backend:1986
- SMITH_BACKEND_ENDPOINT=${SMITH_BACKEND_ENDPOINT:-http://langchain-backend:1984}
- LANGSMITH_LICENSE_KEY=${LANGSMITH_LICENSE_KEY}
- LOG_LEVEL=${LOG_LEVEL:-info}
- AUTH_TYPE=${AUTH_TYPE:-none}
- OAUTH_CLIENT_ID=${OAUTH_CLIENT_ID}
- OAUTH_CLIENT_SECRET=${OAUTH_CLIENT_SECRET}
- OAUTH_ISSUER_URL=${OAUTH_ISSUER_URL}
- API_KEY_SALT=${API_KEY_SALT}
- X_SERVICE_AUTH_JWT_SECRET=${API_KEY_SALT}
- POSTGRES_DATABASE_URI=${POSTGRES_DATABASE_URI:-postgres:postgres@langchain-db:5432/postgres}
- REDIS_DATABASE_URI=${REDIS_DATABASE_URI:-redis://langchain-redis:6379}
- CLICKHOUSE_HOST=${CLICKHOUSE_HOST:-langchain-clickhouse}
- CLICKHOUSE_USER=${CLICKHOUSE_USER:-default}
- CLICKHOUSE_PASSWORD=${CLICKHOUSE_PASSWORD:-password}
- CLICKHOUSE_DB=${CLICKHOUSE_DB:-default}
- CLICKHOUSE_PORT=${CLICKHOUSE_PORT:-8123}
- CLICKHOUSE_TLS=${CLICKHOUSE_TLS:-false}
- FF_ORG_CREATION_DISABLED=${ORG_CREATION_DISABLED:-false}
- FF_TRACE_TIERS_ENABLED=${TTL_ENABLED:-true}
- FF_UPGRADE_TRACE_TIER_ENABLED=${TTL_ENABLED:-true}
- FF_S3_STORAGE_ENABLED=${BLOB_STORAGE_ENABLED:-false}
- S3_BUCKET_NAME=${BLOB_STORAGE_BUCKET_NAME:-langsmith-s3-assets}
- S3_RUN_MANIFEST_BUCKET_NAME=${BLOB_STORAGE_BUCKET_NAME:-langsmith-s3-assets}
- S3_API_URL=${BLOB_STORAGE_API_URL:-https://s3.us-west-2.amazonaws.com}
- S3_ACCESS_KEY=${BLOB_STORAGE_ACCESS_KEY}
- S3_ACCESS_KEY_SECRET=${BLOB_STORAGE_ACCESS_KEY_SECRET}
- FF_CH_SEARCH_ENABLED=${CH_SEARCH_ENABLED:-true}
- BASIC_AUTH_ENABLED=${BASIC_AUTH_ENABLED:-false}
- BASIC_AUTH_JWT_SECRET=${BASIC_AUTH_JWT_SECRET}
- INITIAL_ORG_ADMIN_EMAIL=${INITIAL_ORG_ADMIN_EMAIL}
- INITIAL_ORG_ADMIN_PASSWORD=${INITIAL_ORG_ADMIN_PASSWORD}
- TRACE_TIER_TTL_DURATION_SEC_MAP=${TRACE_TIER_TTL_DURATION_SEC_MAP}
ports:
- 1984:1984
depends_on:
langchain-db:
condition: service_healthy
langchain-redis:
condition: service_healthy
clickhouse-setup:
condition: service_completed_successfully
postgres-setup:
condition: service_completed_successfully
restart: always
langchain-platform-backend:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-go-backend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
environment:
- PORT=1986
- LANGCHAIN_ENV=local_docker
- LANGSMITH_URL=${LANGSMITH_URL:-http://langchain-frontend:1980}
- SMITH_BACKEND_ENDPOINT=${SMITH_BACKEND_ENDPOINT:-http://langchain-backend:1984}
- LANGSMITH_LICENSE_KEY=${LANGSMITH_LICENSE_KEY}
- LOG_LEVEL=${LOG_LEVEL:-warning}
- AUTH_TYPE=${AUTH_TYPE:-none}
- OAUTH_CLIENT_ID=${OAUTH_CLIENT_ID}
- OAUTH_CLIENT_SECRET=${OAUTH_CLIENT_SECRET}
- OAUTH_ISSUER_URL=${OAUTH_ISSUER_URL}
- API_KEY_SALT=${API_KEY_SALT}
- X_SERVICE_AUTH_JWT_SECRET=${API_KEY_SALT}
- POSTGRES_DATABASE_URI=${POSTGRES_DATABASE_URI:-postgres:postgres@langchain-db:5432/postgres}
- REDIS_DATABASE_URI=${REDIS_DATABASE_URI:-redis://langchain-redis:6379}
- BASIC_AUTH_ENABLED=${BASIC_AUTH_ENABLED:-false}
- BASIC_AUTH_JWT_SECRET=${BASIC_AUTH_JWT_SECRET}
- CLICKHOUSE_HOST=${CLICKHOUSE_HOST:-langchain-clickhouse}
- CLICKHOUSE_USER=${CLICKHOUSE_USER:-default}
- CLICKHOUSE_PASSWORD=${CLICKHOUSE_PASSWORD:-password}
- CLICKHOUSE_DB=${CLICKHOUSE_DB:-default}
- CLICKHOUSE_PORT=${CLICKHOUSE_PORT:-8123}
- CLICKHOUSE_TLS=${CLICKHOUSE_TLS:-false}
- FF_ORG_CREATION_DISABLED=${ORG_CREATION_DISABLED:-false}
- FF_TRACE_TIERS_ENABLED=${TTL_ENABLED:-true}
- FF_UPGRADE_TRACE_TIER_ENABLED=${TTL_ENABLED:-true}
- FF_S3_STORAGE_ENABLED=${BLOB_STORAGE_ENABLED:-false}
- S3_BUCKET_NAME=${BLOB_STORAGE_BUCKET_NAME:-langsmith-s3-assets}
- S3_RUN_MANIFEST_BUCKET_NAME=${BLOB_STORAGE_BUCKET_NAME:-langsmith-s3-assets}
- S3_API_URL=${BLOB_STORAGE_API_URL:-https://s3.us-west-2.amazonaws.com}
- S3_ACCESS_KEY=${BLOB_STORAGE_ACCESS_KEY}
- S3_ACCESS_KEY_SECRET=${BLOB_STORAGE_ACCESS_KEY_SECRET}
- FF_CH_SEARCH_ENABLED=${CH_SEARCH_ENABLED:-true}
- TRACE_TIER_TTL_DURATION_SEC_MAP=${TRACE_TIER_TTL_DURATION_SEC_MAP}
ports:
- 1986:1986
depends_on:
langchain-db:
condition: service_healthy
langchain-redis:
condition: service_healthy
clickhouse-setup:
condition: service_completed_successfully
postgres-setup:
condition: service_completed_successfully
restart: always
langchain-queue:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-backend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
environment:
- LANGCHAIN_ENV=local_docker
- GO_ENDPOINT=http://langchain-platform-backend:1986
- SMITH_BACKEND_ENDPOINT=http://langchain-backend:1984
- LANGSMITH_LICENSE_KEY=${LANGSMITH_LICENSE_KEY}
- LOG_LEVEL=${LOG_LEVEL:-info}
- AUTH_TYPE=${AUTH_TYPE:-none}
- OAUTH_CLIENT_ID=${OAUTH_CLIENT_ID}
- OAUTH_ISSUER_URL=${OAUTH_ISSUER_URL}
- API_KEY_SALT=${API_KEY_SALT}
- X_SERVICE_AUTH_JWT_SECRET=${API_KEY_SALT}
- POSTGRES_DATABASE_URI=${POSTGRES_DATABASE_URI:-postgres:postgres@langchain-db:5432/postgres}
- REDIS_DATABASE_URI=${REDIS_DATABASE_URI:-redis://langchain-redis:6379}
- CLICKHOUSE_HOST=${CLICKHOUSE_HOST:-langchain-clickhouse}
- CLICKHOUSE_USER=${CLICKHOUSE_USER:-default}
- CLICKHOUSE_PASSWORD=${CLICKHOUSE_PASSWORD:-password}
- CLICKHOUSE_DB=${CLICKHOUSE_DB:-default}
- CLICKHOUSE_PORT=${CLICKHOUSE_PORT:-8123}
- CLICKHOUSE_TLS=${CLICKHOUSE_TLS:-false}
- FF_ORG_CREATION_DISABLED=${ORG_CREATION_DISABLED:-false}
- FF_TRACE_TIERS_ENABLED=${TTL_ENABLED:-true}
- FF_UPGRADE_TRACE_TIER_ENABLED=${TTL_ENABLED:-true}
- FF_S3_STORAGE_ENABLED=${BLOB_STORAGE_ENABLED:-false}
- S3_BUCKET_NAME=${BLOB_STORAGE_BUCKET_NAME:-langsmith-s3-assets}
- S3_RUN_MANIFEST_BUCKET_NAME=${BLOB_STORAGE_BUCKET_NAME:-langsmith-s3-assets}
- S3_API_URL=${BLOB_STORAGE_API_URL:-https://s3.us-west-2.amazonaws.com}
- S3_ACCESS_KEY=${BLOB_STORAGE_ACCESS_KEY}
- S3_ACCESS_KEY_SECRET=${BLOB_STORAGE_ACCESS_KEY_SECRET}
- FF_CH_SEARCH_ENABLED=${CH_SEARCH_ENABLED:-true}
- BASIC_AUTH_ENABLED=${BASIC_AUTH_ENABLED:-false}
- BASIC_AUTH_JWT_SECRET=${BASIC_AUTH_JWT_SECRET}
- TRACE_TIER_TTL_DURATION_SEC_MAP=${TRACE_TIER_TTL_DURATION_SEC_MAP}
command:
- "saq"
- "app.workers.queues.single_queue_worker.settings"
- "--quiet"
depends_on:
langchain-db:
condition: service_healthy
langchain-redis:
condition: service_healthy
clickhouse-setup:
condition: service_completed_successfully
postgres-setup:
condition: service_completed_successfully
restart: always
langchain-db:
image: ${_REGISTRY:-docker.io}/postgres:14.7
command:
[
"postgres",
"-c",
"log_min_messages=WARNING",
"-c",
"client_min_messages=WARNING",
]
environment:
- POSTGRES_PASSWORD=postgres
- POSTGRES_USER=postgres
- POSTGRES_DB=postgres
volumes:
- langchain-db-data:/var/lib/postgresql/data
ports:
- 5433:5432
healthcheck:
test: ["CMD", "pg_isready", "-U", "postgres"]
interval: 2s
timeout: 2s
retries: 30
langchain-redis:
image: ${_REGISTRY:-docker.io}/redis:7
ports:
- 63791:6379
volumes:
- langchain-redis-data:/data
healthcheck:
test: ["CMD", "redis-cli", "ping"]
interval: 2s
timeout: 2s
retries: 30
langchain-clickhouse:
image: ${_REGISTRY:-docker.io}/clickhouse/clickhouse-server:24.8
user: "101:101"
restart: always
environment:
- CLICKHOUSE_DB=${CLICKHOUSE_DB:-default}
- CLICKHOUSE_USER=${CLICKHOUSE_USER:-default}
- CLICKHOUSE_PASSWORD=${CLICKHOUSE_PASSWORD:-password}
volumes:
- langchain-clickhouse-data:/var/lib/clickhouse
- ./users.xml:/etc/clickhouse-server/users.d/users.xml
ports:
- 8124:8123
- 9001:9000
healthcheck:
test: ["CMD", "clickhouse-client", "--query", "SELECT 1"]
interval: 2s
timeout: 2s
retries: 30
clickhouse-setup:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-backend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
depends_on:
langchain-clickhouse:
condition: service_healthy
restart: "on-failure:10"
environment:
- CLICKHOUSE_HOST=${CLICKHOUSE_HOST:-langchain-clickhouse}
- CLICKHOUSE_USER=${CLICKHOUSE_USER:-default}
- CLICKHOUSE_PASSWORD=${CLICKHOUSE_PASSWORD:-password}
- CLICKHOUSE_DB=${CLICKHOUSE_DB:-default}
- CLICKHOUSE_PORT=${CLICKHOUSE_PORT:-8123}
- CLICKHOUSE_NATIVE_PORT=${CLICKHOUSE_NATIVE_PORT:-9000}
- CLICKHOUSE_TLS=${CLICKHOUSE_TLS:-false}
command:
[
"bash",
"scripts/wait_for_clickhouse_and_migrate.sh"
]
postgres-setup:
image: ${_REGISTRY:-docker.io}/langchain/langsmith-backend:${_LANGSMITH_IMAGE_VERSION:-0.11.4}
depends_on:
langchain-db:
condition: service_healthy
environment:
- LANGCHAIN_ENV=local_docker
- LANGSMITH_LICENSE_KEY=${LANGSMITH_LICENSE_KEY}
- LOG_LEVEL=${LOG_LEVEL:-warning}
- AUTH_TYPE=${AUTH_TYPE:-none}
- OAUTH_CLIENT_ID=${OAUTH_CLIENT_ID}
- OAUTH_ISSUER_URL=${OAUTH_ISSUER_URL}
- API_KEY_SALT=${API_KEY_SALT}
- POSTGRES_DATABASE_URI=${POSTGRES_DATABASE_URI:-postgres:postgres@langchain-db:5432/postgres}
- REDIS_DATABASE_URI=${REDIS_DATABASE_URI:-redis://langchain-redis:6379}
- MAX_ASYNC_JOBS_PER_WORKER=${MAX_ASYNC_JOBS_PER_WORKER:-10}
- ASYNCPG_POOL_MAX_SIZE=${ASYNCPG_POOL_MAX_SIZE:-3}
- CLICKHOUSE_HOST=${CLICKHOUSE_HOST:-langchain-clickhouse}
- CLICKHOUSE_USER=${CLICKHOUSE_USER:-default}
- CLICKHOUSE_PASSWORD=${CLICKHOUSE_PASSWORD:-password}
- CLICKHOUSE_DB=${CLICKHOUSE_DB:-default}
- CLICKHOUSE_PORT=${CLICKHOUSE_PORT:-8123}
- CLICKHOUSE_NATIVE_PORT=${CLICKHOUSE_NATIVE_PORT:-9000}
- CLICKHOUSE_TLS=${CLICKHOUSE_TLS:-false}
restart: "on-failure:10"
command:
[
"bash",
"-c",
"alembic upgrade head",
]
volumes:
langchain-db-data:
langchain-redis-data:
langchain-clickhouse-data:
.env
# Don't change this file. Instead, copy it to .env and change the values there. The default values will work out of the box as long as you provide your license key.
_REGISTRY=docker.io # Change to your desired Docker registry if you are using a private registry. Otherwise, leave it as is
_LANGSMITH_IMAGE_VERSION=0.11.3 # Change to the desired Langsmith image version
LANGSMITH_LICENSE_KEY= # Change to your Langsmith license key
AUTH_TYPE=none # Set to oauth if you want to use OAuth2.0 with PKCE. Set to mixed for basic auth or OAuth2.0 with OAuth2.0 client secret
OAUTH_CLIENT_ID="" # Required if AUTH_TYPE=oauth or mixed with OAuth2.0 with OAuth2.0 client secret
OAUTH_ISSUER_URL="" # Required if AUTH_TYPE=oauth or mixed with OAuth2.0 with OAuth2.0 client secret (https://your-issuer-url)
OAUTH_CLIENT_SECRET="" # Required if AUTH_TYPE=none with OAuth2.0 with OAuth2.0 client secret
LANGSMITH_URL=http://localhost:1980 # Change to your hosted Langsmith URL. Required if AUTH_TYPE=none with OAuth2.0 client secret
API_KEY_SALT=langsmith-secret-key-2024-random-salt-value-for-testing # Change to your desired API key salt. Can be any random value. Must be set if AUTH_TYPE=oauth
POSTGRES_DATABASE_URI=postgres:postgres@langchain-db:5432/postgres # Change to your database URI if using external postgres. Otherwise, leave it as is
REDIS_DATABASE_URI=redis://langchain-redis:6379 # Change to your Redis URI if using external Redis. Otherwise, leave it as is
LOG_LEVEL=info # Change to your desired log level
MAX_ASYNC_JOBS_PER_WORKER=10 # Change to your desired maximum async jobs per worker. We recommend 10/suggest spinning up more replicas of the queue worker if you need more throughput
ASYNCPG_POOL_MAX_SIZE=3 # Change the PG pool size based off your pg instance/requirements.
CLICKHOUSE_HOST=langchain-clickhouse # Change to your Clickhouse host if using external Clickhouse. Otherwise, leave it as is
CLICKHOUSE_USER=default # Change to your Clickhouse user if needed
CLICKHOUSE_DB=default # Change to your Clickhouse database if needed
CLICKHOUSE_PORT=8123 # Change to your Clickhouse port if needed
CLICKHOUSE_TLS=false # Change to true if you are using TLS to connect to Clickhouse. Otherwise, leave it as is
CLICKHOUSE_CLUSTER= # Change to your Clickhouse cluster if using a replicated cluster. Otherwise, leave it as is
CLICKHOUSE_PASSWORD=password # Change to your Clickhouse password if needed
CLICKHOUSE_NATIVE_PORT=9000 # Change to your Clickhouse native port if needed
ORG_CREATION_DISABLED=false # Set to true if you want to disable org creation
WORKSPACE_SCOPE_ORG_INVITES_ENABLED=false # Set to true if you want to disable workspace scope org invites
PERSONAL_ORGS_DISABLED=false # Set to true if you want to disable personal orgs
TTL_ENABLED=true # Set to true if you want to enable TTL for your data
TRACE_TIER_TTL_DURATION_SEC_MAP='{"longlived":34560000,"shortlived":1209600}'
BLOB_STORAGE_ENABLED=false # Set to true if you want to enable blob storage
BLOB_STORAGE_BUCKET_NAME=langsmith-blob-storage # Change to your desired blob storage bucket name
BLOB_STORAGE_API_URL=https://s3.us-west-2.amazonaws.com # Change to your desired blob storage API URL
BLOB_STORAGE_ACCESS_KEY=your-access-key # Change to your desired blob storage access key
BLOB_STORAGE_ACCESS_KEY_SECRET=your-access-key-secret # Change to your desired blob storage access key secret
CH_SEARCH_ENABLED=true # Set to false if you do not want to store tokenized inputs/outputs in clickhouse
BASIC_AUTH_ENABLED=false # Set to true if you want to enable basic auth
BASIC_AUTH_JWT_SECRET=langsmith-jwt-secret-key-2024-random-value-for-testing # Change to your desired basic auth JWT secret
INITIAL_ORG_ADMIN_EMAIL=admin@example.com # Change to your desired initial org admin email. Only used if BASIC_AUTH_ENABLED=false
INITIAL_ORG_ADMIN_PASSWORD=AdminPassword123! # Change to your desired initial org admin password. Needs to be at least 12 characters long, contain at least one lowercase, one uppercase, and one special character. Only used if BASIC_AUTH_ENABLED=falseLANGSMITH_API_KEY=sk-test-key-2024
LANGCHAIN_TRACING_V2=false
DISABLE_AUTH=true
PUBLIC_ACCESS=true
ALLOW_ANONYMOUS_ACCESS=true
LANGSMITH_PUBLIC=true
SKIP_AUTH=true
AUTH_ENABLED=false
LANGSMITH_SELF_HOSTED=true
TENANT_CREATION_DISABLED=false
ENABLE_FREE_TIER=true
LANGSMITH_DEFAULT_TENANT=default
LANGSMITH_DEFAULT_ORG=default-org
AUTO_CREATE_TENANT=true
users.xml
1
1
1
1
default
1
2000000
0
1
0
25
0
0
启动
docker-compose up
访问下面的界面
LANGSMITH_LICENSE_KEY 这个目前是需要联系 官网的销售团队 Talk to our team 点击链接[[Talk to our team]去填写信息
三 k8s 部署 langsmith
3.1 minikube安装 k8s
minikube 是一个工具， 能让你在本地运行 Kubernetes。 minikube 在你的个人计算机（包括 Windows、macOS 和 Linux PC）上运行一个一体化（all-in-one） 或多节点的本地 Kubernetes 集群，以便你来尝试 Kubernetes 或者开展每天的开发工作。
安装命令
winget install Kubernetes.minikube
以管理员身份启动一个cmd
3.2 输入 minikube start 启动群集
3.3 安装 kubectl
下载
curl.exe -LO "https://dl.k8s.io/release/v1.34.0/bin/windows/amd64/kubectl.exe"
放到D:kubectl， 然后设置环境变量
查看你 kubectl 的版本
3.4 安装 helm
choco install kubernetes-helm
生成 Api Key Salt
C:UsersAdministrator>openssl rand -base64 32
dXZFfZy9c0m16eIi9Vyqe1IuaXuCVDNbsmU/X7Sg5FE=
生成JWT Secret
C:UsersAdministrator>openssl rand -base64 32
8DvHV/Am3DVk6eWMiQxiLFZd0OTiCh2LgZ5IX2Wv658=
添加 Helm 仓库
helm repo add langchain https://langchain-ai.github.io/helm/
查找最新版本
helm search repo langchain/langsmith --versions
langchain/langsmith 0.12.3 0.12.10 Helm chart to deploy the langsmith application ...
langchain/langsmith 0.12.2 0.12.10 Helm chart to deploy the langsmith application ...
langchain/langsmith 0.12.1 0.12.10 Helm chart to deploy the langsmith application ...
langchain/langsmith 0.12.0 0.12.10 Helm chart to deploy the langsmith application ...
langchain/langsmith 0.11.29 0.12.4 Helm chart to deploy the langsmith application ...
langchain/langsmith 0.11.28 0.12.4 Helm chart to deploy the langsmith application ...
新建一个 langsmith_config.yaml
config:
langsmithLicenseKey: "" 这里需要自己申请的key
apiKeySalt: "dXZFfZy9c0m16eIi9Vyqe1IuaXuCVDNbsmU/X7Sg5FE="
authType: mixed
basicAuth:
enabled: true
initialOrgAdminEmail: "admin@example.com"
initialOrgAdminPassword: "YourSecurePassword123!"
jwtSecret: "8DvHV/Am3DVk6eWMiQxiLFZd0OTiCh2LgZ5IX2Wv658="
为 Minikube 启用默认存储类
安装/升级 LangSmith
helm upgrade -i langsmith langchain/langsmith --values langsmith_config.yaml --version 0.12.2 --debug
备注 如果失败了，ctrl+c结束
先需要执行kubectl delete pvc -l app.kubernetes.io/instance=langsmith 进行删除 ，然后清理当前失败的部署：helm uninstall langsmith
检查部署状态
kubectl get pods -w
langsmith-backend-6ff46c99c4-wz22d 1/1 Running 0 3h2m
langsmith-frontend-6bbb94c5df-8xrlr 1/1 Running 0 3h2m
langsmith-hub-backend-5cc68c888c-vppjj 1/1 Running 0 3h2m
langsmith-playground-6d95fd8dc6-x2d9b 1/1 Running 0 3h2m
langsmith-postgres-0 1/1 Running 0 9h
langsmith-queue-5898b9d566-tv6q8 1/1 Running 0 3h2m
langsmith-redis-0 1/1 Running 0 9h
kubectl get services
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
langsmith-backend ClusterIP 172.20.140.77 1984/TCP 35h
langsmith-frontend LoadBalancer 172.20.253.251 80:31591/TCP 35h
langsmith-hub-backend ClusterIP 172.20.112.234 1985/TCP 35h
langsmith-playground ClusterIP 172.20.153.194 3001/TCP 9h
langsmith-postgres ClusterIP 172.20.244.82 5432/TCP 35h
langsmith-redis ClusterIP 172.20.81.217 6379/TCP 35h
浏览器访问 langsmith-frontend界面
LANGSMITH_LICENSE_KEY 这个目前是需要联系 官网的销售团队 Talk to our team 点击链接[Talk to our team]去填写信息
总结： 独立服务器(standalone server) 对于日常的开发是足够的， 还可以搭云端的langsmith来使用。
本地的langsmith和langsmith with deployment 不建议本地安装，第一个原因是机器的硬件要求：docker安装建议4 vCPUs, 16GB Memory available on your machine；k8s安装建议Recommended: At least 16 vCPUs, 64GB Memory available； 第二个原因 是不同机器环境配置，会出现的错误会很多，按照步骤也会遇到新的问题，需要自己根据实际情况去解决。第三个原因是需要企业级， 需要申请 LANGSMITH_LICENSE_KEY 认证许可。 langsmith这个适合企业 公司按照，个人开发没必要那么费劲折腾。
说真的，这两年看着身边一个个搞Java、C++、前端、数据、架构的开始卷大模型，挺唏嘘的。大家最开始都是写接口、搞Spring Boot、连数据库、配Redis，稳稳当当过日子。
结果GPT、DeepSeek火了之后，整条线上的人都开始有点慌了，大家都在想：“我是不是要学大模型，不然这饭碗还能保多久？”
我先给出最直接的答案：一定要把现有的技术和大模型结合起来，而不是抛弃你们现有技术！掌握AI能力的Java工程师比纯Java岗要吃香的多。
即使现在裁员、降薪、团队解散的比比皆是……但后续的趋势一定是AI应用落地！大模型方向才是实现职业升级、提升薪资待遇的绝佳机遇！
这绝非空谈。数据说话
2025年的最后一个月，脉脉高聘发布了《2025年度人才迁徙报告》，披露了2025年前10个月的招聘市场现状。
AI领域的人才需求呈现出极为迫切的“井喷”态势
2025年前10个月，新发AI岗位量同比增长543%，9月单月同比增幅超11倍。同时，在薪资方面，AI领域也显著领先。其中，月薪排名前20的高薪岗位平均月薪均超过6万元，而这些席位大部分被AI研发岗占据。
与此相对应，市场为AI人才支付了显著的溢价：算法工程师中，专攻AIGC方向的岗位平均薪资较普通算法工程师高出近18%；产品经理岗位中，AI方向的产品经理薪资也领先约20%。
当你意识到“技术+AI”是个人突围的最佳路径时，整个就业市场的数据也印证了同一个事实：AI大模型正成为高薪机会的最大源头。
最后
我在一线科技企业深耕十二载，见证过太多因技术卡位而跃迁的案例。那些率先拥抱 AI 的同事，早已在效率与薪资上形成代际优势，我意识到有很多经验和知识值得分享给大家，也可以通过我们的能力和经验解答大家在大模型的学习中的很多困惑。
我整理出这套 AI 大模型突围资料包【允许白嫖】：
-
✅从入门到精通的全套视频教程
-
✅AI大模型学习路线图（0基础到项目实战仅需90天）
-
✅大模型书籍与技术文档PDF
-
✅各大厂大模型面试题目详解
-
✅640套AI大模型报告合集
-
✅大模型入门实战训练
这份完整版的大模型 AI 学习和面试资料已经上传CSDN，朋友们如果需要可以微信扫描下方CSDN官方认证二维码免费领取【保证100%免费】
①从入门到精通的全套视频教程
包含提示词工程、RAG、Agent等技术点
② AI大模型学习路线图（0基础到项目实战仅需90天）
全过程AI大模型学习路线
③学习电子书籍和技术文档
市面上的大模型书籍确实太多了，这些是我精选出来的
④各大厂大模型面试题目详解
⑤640套AI大模型报告合集
⑥大模型入门实战训练
👉获取方式：
有需要的小伙伴，可以保存图片到wx扫描二v码免费领取【保证100%免费】🆓
