After executing a large number of tasks, the CPU of the worker container and api container remains highly occupied.

[dyyTT]

Self Checks

• This is only for bug report, if you would like to ask a question, please head to Discussions.
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• Please do not modify this template :) and fill in all the required fields.

Dify version

0.13.1

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

After executing a large number of tasks, the CPU of the worker container and api container remains highly occupied.

✔️ Expected Behavior

No response

❌ Actual Behavior

No response

[dosubot]

Hey @dyyTT! I'm here to help you with any bugs, questions, or contributions you have for Dify. Just tag me if you need assistance!

High CPU usage in Dify's worker and API containers can be caused by several factors, especially when processing a large number of tasks. Here are some potential solutions and optimizations you can consider:

1. Adjust Worker and API Configuration:

   • Increase the number of Celery workers (CELERY_WORKER_AMOUNT) to distribute the load more evenly.
   • Adjust the SERVER_WORKER_AMOUNT to better utilize available CPU cores ^[1].

2. Optimize Task Execution:

   • Review the tasks being executed to ensure they are optimized for performance. Consider breaking down large tasks into smaller, more manageable ones.

3. Rate Limiting and Throttling:

   • Implement rate limiting to prevent overwhelming the system with too many simultaneous requests, which can lead to high CPU usage ^[2].

4. Resource Allocation:

   • Ensure that your Docker setup is configured to allocate sufficient resources (CPU and memory) to the containers. This can be done by adjusting Docker's resource limits.

5. Monitor and Analyze:

   • Use monitoring tools to analyze the performance and identify any bottlenecks or specific tasks that are consuming excessive resources.

These steps should help in managing and reducing the CPU usage of your Dify containers.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[dyyTT]

嘿 ！我在这里帮助您解决您对 Dify 的任何错误、问题或贡献。如果您需要帮助，请标记我！

Dify 的工作程序和 API 容器中的高 CPU 使用率可能由多种因素引起，尤其是在处理大量任务时。以下是您可以考虑的一些潜在解决方案和优化：

1. 调整 Worker 和 API 配置：

   • 增加 Celery 工作线程 （） 的数量以更均匀地分配负载。CELERY_WORKER_AMOUNT
   • 调整以更好地利用可用的 CPU 内核SERVER_WORKER_AMOUNT[1].

2. 优化任务执行：

   • 查看正在执行的任务，以确保它们针对性能进行了优化。考虑将大型任务分解为更小、更易于管理的任务。

3. 速率限制和限制：

   • 实施速率限制，以防止系统因同时请求过多而不堪重负，从而导致 CPU 使用率过高[2].

4. 资源分配：

   • 确保您的 Docker 设置配置为为容器分配足够的资源（CPU 和内存）。这可以通过调整 Docker 的资源限制来完成。

5. 监控和分析：

   • 使用监控工具分析性能并识别任何瓶颈或消耗过多资源的特定任务。

这些步骤应该有助于管理和降低 Dify 容器的 CPU 使用率。

要继续与 Dosu 交谈，请提及 。

帮助 Dosu 学习！给它反馈：伟大的回应 |无关紧要的答案 |不正确的来源 |太冗长 |幻觉 |其他

But why does the CPU usage remain high even after all the tasks have been completed?

[dosubot]

High CPU usage after task completion in Dify could be due to several factors:

1. Lingering Processes: Dify's task queue management might have lingering processes if the stop conditions are not met or if the stop flag in Redis is not set correctly. Ensure that the APP_MAX_EXECUTION_TIME is appropriately configured and that the stop flag is being set and checked properly ^[1].

2. Resource Allocation: Check if the Docker containers are configured with appropriate resource limits. Insufficient resource allocation can cause processes to remain active longer than necessary.

3. Monitoring and Logging: Although specific built-in monitoring tools were not identified, implementing external monitoring solutions can help track resource usage and identify processes that continue to consume resources.

4. Celery Worker Management: Ensure that Celery workers are configured to terminate correctly after task completion. Misconfigurations in worker lifecycle management can lead to high CPU usage.

Review these areas to identify and address the cause of high CPU usage after task completion.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[dyyTT]

Why does the CPU usage of Gunicorn remain high even after a large number of API requests have ended?
**

**
@dosu

[dosubot]

High CPU usage by Gunicorn after API requests have ended can be due to several factors:

1. Worker Processes: Gunicorn's worker processes might not be terminating correctly after handling requests. This can happen if the GUNICORN_TIMEOUT is not set appropriately, causing workers to remain active longer than necessary. Ensure that the timeout is configured to match your application's needs ^[1].

2. Concurrency Handling: The SERVER_WORKER_CLASS is set to gevent by default, which is efficient for handling many simultaneous connections. However, if the number of connections or requests is not managed properly, it can lead to high CPU usage. Consider adjusting the SERVER_WORKER_CONNECTIONS to better handle your load ^[2].

3. Resource Management: Check if the number of workers (SERVER_WORKER_AMOUNT) is optimal for your CPU core count. The default calculation is CPU core count x 2 + 1, which might need adjustment based on your specific workload and server capabilities ^[1].

4. Background Tasks: If there are background tasks or long-running requests, they might be keeping the workers busy. Review your application's logic to ensure that tasks are being handled efficiently and that any unnecessary processing is minimized.

Review these areas to identify potential causes and adjust configurations to reduce CPU usage.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}