dataset parsing task has been in queued for 600s+

[jiangsanyin]

Describe your problem

I deployed ragflow 0.15.1 with ragflow/docker/docker-compose.yml in physical machine which Ubuntu20.04 LTS OS was installed, and this machine has 2 NVIDIA A40 GPU.

1. I created a file parsing task with file "食品安全法.pdf", and it worked well.
2. Then I directly redeployed ragflow 0.15.1 by running "docker-compose -f docker-compose-gpu.yml up -d" , and all docker containers is fine:

3) Now I try to create a file parsing task with file "乳腺癌相关说明（作为知识库）.pdf", however this task is always in queued

How do I look into the logs and locate this problem?

[EBazarov]

You can check logs of docker container to see if there's anything weird. Example:

docker logs -f ragflow-server

[shinegob]

同样的问题，解决了嘛

[jiangsanyin]

You can check logs of docker container to see if there's anything weird. Example:

docker logs -f ragflow-server

Thanks for your reply. I checked it but found nothing abnormal

[jiangsanyin]

同样的问题，解决了嘛

我将默认的使用all GPUs改成使用1个GPU，然后就没有问题了。（I used a GPU instead and it works fine then）

[yyx002]

同样的问题，解决了嘛

我将默认使用的all GPUs改成使用1个GPU，然后就没有问题了。（我用了 GPU，然后它工作正常）

请问这个从哪里改，我和你遇到的问题一样，任务总是在排队中

[jiangsanyin]

同样的问题，解决了嘛

我将默认使用的all GPUs改成使用1个GPU，然后就没有问题了。（我用了 GPU，然后它工作正常）

请问这个从哪里改，我和你遇到的问题一样，任务总是在排队中

修改docker-compose-gpu.yml 文件的倒数第2行的内容
root@controller01:/opt/code_repos/ragflow/docker# cat docker-compose-gpu.yml
...
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: 1
capabilities: [gpu]

[yyx002]

感谢感谢，我的也成功了，你知道这是为啥嘛，求教

[shinegob]

感谢，尝试了你的方法，好像确实暂时没有出现全部在排队的情况。目前有个新问题，开启raptor后，除了个别docx文档解析成功，绝大多数斗出现The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()的报错问题，这个请问你出现了吗 ------------------------------------------------------------------ 发件人：yyx002 ***@***.***> 发送时间：2025年2月10日(星期一) 15:10 ***@***.***> ***@***.***>; ***@***.***> 主　题：Re: [infiniflow/ragflow] dataset parsing task has been in queued for 600s+ (Issue #4746) 同样的问题，解决了嘛 我将默认使用的all GPUs改成使用1个GPU，然后就没有问题了。（我用了 GPU，然后它工作正常） 请问这个从哪里改，我和你遇到的问题一样，任务总是在排队中 — Reply to this email directly, view it on GitHub <#4746 (comment) >, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKQMSBGZJ3MQQGUE6MUNTF32PBGG5AVCNFSM6AAAAABWS7M6ICVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNBXGEYTSMRSHA >. You are receiving this because you commented.Message ID: ***@***.***>

[shinegob]

感谢，尝试了你的方法，好像确实暂时没有出现全部在排队的情况。目前有个新问题，开启raptor后，除了个别docx文档解析成功，绝大多数斗出现The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()的报错问题，这个请问你出现了吗 ------------------------------------------------------------------ 发件人：jiangsanyin ***@***.***> 发送时间：2025年2月10日(星期一) 15:13 ***@***.***> ***@***.***>; ***@***.***> 主　题：Re: [infiniflow/ragflow] dataset parsing task has been in queued for 600s+ (Issue #4746) 同样的问题，解决了嘛 我将默认使用的all GPUs改成使用1个GPU，然后就没有问题了。（我用了 GPU，然后它工作正常） 请问这个从哪里改，我和你遇到的问题一样，任务总是在排队中 ***@***.***:/opt/code_repos/ragflow/docker# cat docker-compose-gpu.yml However, you are welcome to file a pull request to improve it. include: * ./docker-compose-base.yml services: ragflow: depends_on: mysql: condition: service_healthy image: ${RAGFLOW_IMAGE} container_name: ragflow-server ports: - ${SVR_HTTP_PORT}:9380 - 80:80 - 444:443 volumes: - ./ragflow-logs:/ragflow/logs - ./nginx/ragflow.conf:/etc/nginx/conf.d/ragflow.conf - ./nginx/proxy.conf:/etc/nginx/proxy.conf - ./nginx/nginx.conf:/etc/nginx/nginx.conf env_file: .env environment: - TZ=${TIMEZONE} - HF_ENDPOINT=${HF_ENDPOINT} - MACOS=${MACOS} networks: - ragflow restart: on-failure # https://docs.docker.com/engine/daemon/prometheus/#create-a-prometheus-configuration <https://docs.docker.com/engine/daemon/prometheus/#create-a-prometheus-configuration > # If you're using Docker Desktop, the --add-host flag is optional. This flag makes sure that the host's internal IP gets exposed to the Prometheus container. extra_hosts: - "host.docker.internal:host-gateway" deploy: resources: reservations: devices: - driver: nvidia count: 1 ##这里从all改为1 capabilities: [gpu] — Reply to this email directly, view it on GitHub <#4746 (comment) >, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKQMSBGDDMIOMBPXQ4LS3YD2PBGRRAVCNFSM6AAAAABWS7M6ICVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNBXGEZDGMRYGE >. You are receiving this because you commented.Message ID: ***@***.***>

[wyqmath]

The web demo version is also always queued and shows the following:
Task is queued...

[lizhao-8202]

我这通过docker-compose-gpu.yml 启动的镜像，启动过后通过nvidia-smi监控，发现ragflow应用没用到gpu样。

我的docker-compose-gpu.yml配置如下

请问是我开启GPU有问题，还是镜像不是GPU版本的还是？

[lizhao-8202]

目前在做解析的时候发现非常慢，很小的1 个文档都要5，60S。.env配置文件中的MEM_LIMIT已经配置到64G了。机器的CPU是32C的，应该不至于这么慢吧

[jiangsanyin]

我这通过docker-compose-gpu.yml 启动的镜像，启动过后通过nvidia-smi监控，发现ragflow应用没用到gpu样。

我的docker-compose-gpu.yml配置如下

请问是我开启GPU有问题，还是镜像不是GPU版本的还是？

我刚刚也试了下，发现在解析文件过程中，宿主机上没有查看到使用nvidia GPU的进程，而且ragflow-server容器中执行“nvidia-smi”报错“Failed to initialize NVML: Unknown Error”。

感觉目前ragflow 0.15.1使用NVIDIA GPU时有问题

[Cedar0607]

Thanks for sharing the solution!!❤