workflow-local-dev

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Workflow Local Development

工作流本地开发

Quick Reference

快速参考

Key Commands

关键命令

Task	Command
Check pods	`kubectl get pods --context kind-kind`
Restart component	`tilt trigger workflow-<service>`
View component logs	`tilt logs workflow-<service>`

任务	命令
检查Pod状态	`kubectl get pods --context kind-kind`
重启组件	`tilt trigger workflow-<service>`
查看组件日志	`tilt logs workflow-<service>`

Locally Deployed Components

本地部署的组件

These workflow services are deployed locally as shared dependencies:

workflow-catalog

workflow-executions-api

workflow-engine-worker

workflow-consumer

workflow-validator

workflows-worker

standalone-tasks-worker

以下工作流服务作为共享依赖项本地部署：

workflow-catalog

workflow-executions-api

workflow-engine-worker

workflow-consumer

workflow-validator

workflows-worker

standalone-tasks-worker

Utility Scripts

实用脚本

Execute these scripts for common operations:

执行以下脚本完成常见操作：

Check Pod Status

检查Pod状态

bash

bash .cursor/skills/workflow-local-dev/scripts/check-pods.sh

bash

bash .cursor/skills/workflow-local-dev/scripts/check-pods.sh

Restart a Service

重启服务

bash

bash .cursor/skills/workflow-local-dev/scripts/restart-service.sh <service-name>

bash

bash .cursor/skills/workflow-local-dev/scripts/restart-service.sh <service-name>

Example: bash .cursor/skills/workflow-local-dev/scripts/restart-service.sh catalog

示例: bash .cursor/skills/workflow-local-dev/scripts/restart-service.sh catalog

undefined

undefined

Tail Service Logs

跟踪服务日志

bash

bash .cursor/skills/workflow-local-dev/scripts/tail-logs.sh <service-name>

bash

bash .cursor/skills/workflow-local-dev/scripts/tail-logs.sh <service-name>

Example: bash .cursor/skills/workflow-local-dev/scripts/tail-logs.sh executions-api

示例: bash .cursor/skills/workflow-local-dev/scripts/tail-logs.sh executions-api

undefined

undefined

Query Database

查询数据库

bash

bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "<SQL query>"

bash

bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "<SQL query>"

Example: bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "SELECT * FROM workflow_engine.workflows ORDER BY created_at DESC LIMIT 5"

示例: bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "SELECT * FROM workflow_engine.workflows ORDER BY created_at DESC LIMIT 5"

---

---

Development Workflow

开发流程

Use this flow primarily when containers fail locally or when you need to debug runtime behavior.

Capture current status:
```
kubectl get pods --context kind-kind
```
Rebuild/restart component:
```
tilt trigger workflow-<service>
```

Watch logs or health:

kubectl logs -f <pod> --context kind-kind

Inspect configuration or data (if needed):

bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "<SQL query>"

主要在容器本地运行失败或需要调试运行时行为时使用此流程。

捕获当前状态:
```
kubectl get pods --context kind-kind
```
重建/重启组件:
```
tilt trigger workflow-<service>
```

监控日志或健康状态:

kubectl logs -f <pod> --context kind-kind

检查配置或数据（如有需要）:

bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "<SQL query>"

Troubleshooting

故障排查

Pod Issues

Pod问题

bash

kubectl get pods --context kind-kind
kubectl describe pod <pod-name> --context kind-kind
kubectl logs -f <pod-name> --context kind-kind

bash

kubectl get pods --context kind-kind
kubectl describe pod <pod-name> --context kind-kind
kubectl logs -f <pod-name> --context kind-kind

Temporal Workflows

Temporal工作流

Open http://localhost:8081 (Temporal UI) and search by workflow ID.

打开 http://localhost:8081 (Temporal UI) 并按工作流ID搜索。

Database State

数据库状态

bash

undefined

bash

undefined

Use the db-query script (runs psql inside the postgres pod)

使用db-query脚本（在postgres Pod内运行psql）

bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "SELECT * FROM workflow_engine.workflows ORDER BY created_at DESC LIMIT 5"

undefined

bash .cursor/skills/workflow-local-dev/scripts/db-query.sh "SELECT * FROM workflow_engine.workflows ORDER BY created_at DESC LIMIT 5"

undefined

Pulsar Messages

Pulsar消息

bash

pulsar-admin topics stats persistent://public/dap/<topic>
pulsar-client consume persistent://public/dap/<topic> -s test-sub -n 10

bash

pulsar-admin topics stats persistent://public/dap/<topic>
pulsar-client consume persistent://public/dap/<topic> -s test-sub -n 10

Using Kubernetes MCP (When Available)

使用Kubernetes MCP（若可用）

If the Kubernetes MCP server is enabled, you can use MCP tools instead of shell commands for a more integrated experience. The MCP defaults to the

kind-kind

context.

如果Kubernetes MCP服务器已启用，您可以使用MCP工具替代Shell命令，获得更集成的体验。MCP默认使用

kind-kind

上下文。

List Pods

列出Pods

Tool: pods_list
Arguments: { "labelSelector": "app=workflow-catalog" }

Or list all pods:

Tool: pods_list
Arguments: {}

工具: pods_list
参数: { "labelSelector": "app=workflow-catalog" }

或列出所有Pods:

工具: pods_list
参数: {}

Get Pod Logs

获取Pod日志

Tool: pods_log
Arguments: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default",
  "tail": 100
}

For previous container logs (after crash):

Tool: pods_log
Arguments: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default",
  "previous": true
}

工具: pods_log
参数: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default",
  "tail": 100
}

对于容器崩溃后的历史日志:

工具: pods_log
参数: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default",
  "previous": true
}

Execute Commands in Pods (e.g., Database Queries)

在Pod中执行命令（例如数据库查询）

Tool: pods_exec
Arguments: {
  "name": "postgres-xxxxx",
  "namespace": "default",
  "command": ["psql", "-U", "postgres", "-d", "temporal", "-c", "SELECT * FROM workflow_engine.workflows ORDER BY created_at DESC LIMIT 5"]
}

工具: pods_exec
参数: {
  "name": "postgres-xxxxx",
  "namespace": "default",
  "command": ["psql", "-U", "postgres", "-d", "temporal", "-c", "SELECT * FROM workflow_engine.workflows ORDER BY created_at DESC LIMIT 5"]
}

Delete Pod (Force Restart)

删除Pod（强制重启）

Tool: pods_delete
Arguments: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default"
}

工具: pods_delete
参数: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default"
}

Get Pod Details

获取Pod详情

Tool: pods_get
Arguments: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default"
}

工具: pods_get
参数: {
  "name": "workflow-catalog-xxxxx",
  "namespace": "default"
}

List Cluster Events (Troubleshooting)

列出集群事件（故障排查）

Tool: events_list
Arguments: { "namespace": "default" }

Note: When using MCP tools, you don't need to specify
context
as it defaults to
kind-kind
. The MCP approach is useful when you want the agent to directly interact with the cluster without spawning shell processes.

工具: events_list
参数: { "namespace": "default" }

注意: 使用MCP工具时，无需指定
context
，因为它默认使用
kind-kind
。当您希望Agent直接与集群交互而无需启动Shell进程时，MCP方法非常有用。

Additional Resources

额外资源

For complete service URLs and infrastructure details, see reference.md

完整的服务URL和基础设施详情，请查看 reference.md