Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Bug] [dolphinscheduler-api] HDFS failed to upload large file #10340

Closed
3 tasks done
github-keep opened this issue Jun 2, 2022 · 31 comments · Fixed by #14268
Closed
3 tasks done

[Bug] [dolphinscheduler-api] HDFS failed to upload large file #10340

github-keep opened this issue Jun 2, 2022 · 31 comments · Fixed by #14268
Assignees
Labels
backend bug Something isn't working priority:high UI ui and front end related
Milestone

Comments

@github-keep
Copy link

Search before asking

  • I had searched in the issues and found no similar issues.

What happened

在配置HDFS之后使用资源中心上传过大的文件(90MB)失败,小文件可以。猜想:不能上传超过HDFS数据块大小(64MB)的文件,查看日志错误信息:没有合适的分解器。
image

What you expected to happen

在之前版本(2.0.5)中,该功能正常可以使用,升级版本3.0.0-alpha,3.0.0-beta-1均不能正常使用。

How to reproduce

集群部署完成后,配置HDFS,使用的Hadoop 集群 NameNode 配置了 HA ,完成之后上传一个128MB(超过HDFS数据块)以上的文件

Anything else

问题每次都出现,相关的日志再安装目录下面的api-server/logs下的最新文件中

Version

3.0.0-beta-1

Are you willing to submit PR?

  • Yes I am willing to submit a PR!

Code of Conduct

@github-keep github-keep added bug Something isn't working Waiting for reply Waiting for reply labels Jun 2, 2022
@github-actions
Copy link

github-actions bot commented Jun 2, 2022

Search before asking

  • I had searched in the issues and found no similar issues.

What happened

After configuring HDFS, uploading too large files (90MB) using the resource center fails, but small files are ok. Guess: cannot upload files larger than the HDFS data block size (64MB), check the log error message: no suitable decomposer.
image

What you expected to happen

In the previous version (2.0.5), this function can be used normally, but the upgraded version 3.0.0-alpha and 3.0.0-beta-1 cannot be used normally.

How to reproduce

After the cluster is deployed, configure HDFS, use the Hadoop cluster NameNode configured with HA, and upload a file over 128MB (more than HDFS data blocks) after completion

Anything else

The problem occurs every time, and the relevant logs are in the latest files under api-server/logs under the installation directory.

Version

3.0.0-beta-1

Are you willing to submit PR?

  • Yes I am willing to submit a PR!

Code of Conduct

@github-actions
Copy link

github-actions bot commented Jun 2, 2022

Thank you for your feedback, we have received your issue, Please wait patiently for a reply.

  • In order for us to understand your request as soon as possible, please provide detailed information、version or pictures.
  • If you haven't received a reply for a long time, you can join our slack and send your question to channel #troubleshooting

@SbloodyS SbloodyS added backend and removed Waiting for reply Waiting for reply labels Jun 2, 2022
@songjianet songjianet added this to the 3.0.0-beta-2 milestone Jun 2, 2022
@davidzollo davidzollo moved this to In Progress in DolphinScheduler Roadmap Jun 6, 2022
@EricGao888
Copy link
Member

EricGao888 commented Jun 9, 2022

I encountered this error, too. May I ask is there any update or potential solutions to this issue?

@fhygh
Copy link

fhygh commented Jun 9, 2022

i upload 200+M file is ok, 3.0.0 beta-1 version
image

@GodWon
Copy link

GodWon commented Jun 20, 2022

我上传 200+M 文件没问题,3.0.0 beta-1 版本 图片

我300M也没报错 传900M的报错了

@SunHeng98
Copy link

I encountered this error, too. May I ask is there any update or potential solutions to this issue?

你好,请问解决了吗?我也遇到这个问题了,好像是上传时间超过15秒就会自动取消。

@devosend devosend modified the milestones: 3.0.0-beta-2, 3.0.0-beta-3 Jul 6, 2022
@yaowj2
Copy link

yaowj2 commented Jul 7, 2022

从debug springboot报的实际异常是org.eclipse.jetty.io.EofException: Early EOF。原因在前端用axios 提交时设置了超时时间为15s,导致在上传过程中前端超时提前中断了, 见下面在ui service.ts的代码片段部分,修改加大timeout时间后上传大文件通过
const baseRequestConfig: AxiosRequestConfig = { baseURL: import.meta.env.MODE === 'development' ? '/dolphinscheduler' : import.meta.env.VITE_APP_PROD_WEB_URL + '/dolphinscheduler', timeout: 15000,

@CriysHot
Copy link

从debug springboot报的实际异常是org.eclipse.jetty.io.EofException: Early EOF。原因在前端用axios 提交时设置了超时时间为15s,导致在上传过程中前端超时提前中断了, 见下面在ui service.ts的代码片段部分,修改加大timeout时间后上传大文件通过
const baseRequestConfig: AxiosRequestConfig = { baseURL: import.meta.env.MODE === 'development' ? '/dolphinscheduler' : import.meta.env.VITE_APP_PROD_WEB_URL + '/dolphinscheduler', timeout: 15000,
Excuse me,i wanna know how to change the conf conveniently?should I compile the code?

@EricGao888
Copy link
Member

related: #10509

@yangjf2019
Copy link

yangjf2019 commented Jul 12, 2022

I'm having the same problem, what specific file configuration needs to be changed, please? The service.ts file was not found. @CriysHot @yaowj2

The file url is https://github.com/apache/dolphinscheduler/blob/dev/dolphinscheduler-ui/src/service/service.ts.

@yangjf2019
Copy link

yangjf2019 commented Jul 12, 2022

Solved, the following 4 files need to be modified.

/api-server/ui/assets/service.766f4632.js
/api-server/ui/assets/service.766f4632.js.gz

/ui/assets/service.766f4632.js
/ui/assets/service.766f4632.js.gz

and find the following configuration baseURL:"/dolphinscheduler",timeout:15e3,then modify 15e3 to 15e5 or greater.

@EricGao888
Copy link
Member

Solved, the following 4 files need to be modified.

/api-server/ui/assets/service.766f4632.js
/api-server/ui/assets/service.766f4632.js.gz

/ui/assets/service.766f4632.js
/ui/assets/service.766f4632.js.gz

and find the following configuration baseURL:"/dolphinscheduler",timeout:15e3,then modify 15e3 to 15e5 or greater.

@yangjf2019 Great job! Would you like to submit a PR to fix this?

@yangjf2019
Copy link

Thanks, I can try.

@yangjf2019
Copy link

Hi, @EricGao888 is it possible to increase the value of this parameter so that it becomes 30 minutes?

@yangjf2019
Copy link

In general, and frankly, it's not recommended to use dolphinscheduler to upload files that are too big!

@EricGao888
Copy link
Member

Hi, @EricGao888 is it possible to increase the value of this parameter so that it becomes 30 minutes?

May I ask whether it is possible to make it configurable for users?

@yangjf2019
Copy link

Yes, I think it should be done too, please let me take another look, thanks.

@zhongjiajie
Copy link
Member

zhongjiajie commented Jul 28, 2022

Thanks, I can try.

Do you still want to submit pull request to fix this? @yangjf2019

@xpf-demo
Copy link

我改了那四个js文件,,为啥还是不行啊,有大佬有相同的情况吗

@github-keep
Copy link
Author

我改了那四个js文件,,为啥还是不行啊,有大佬有相同的情况吗

The larger the file, the longer the upload time. It is recommended to set the time value larger

@xpf-demo
Copy link

我改了那四个js文件,,为啥还是不行啊,有大佬有同样的情况吗

文件越大,上传时间越长。建议将时间值设置得大一些

谢谢,我的问题已经解决了。我后面修改了js文件之后,清空了下浏览器的缓存之后,就能正常上传了,后台也没有报错了

@EricGao888 EricGao888 added the UI ui and front end related label Aug 23, 2022
@EricGao888
Copy link
Member

Yes, I think it should be done too, please let me take another look, thanks.

Hello @yangjf2019, may I ask whether you are still working on this issue? We have received feedback from many users that they get blocked by this issue. IMHO we could have a hot-fix for it at the first step, simply increasing the threshold. Then we could move it further and make it configurable for users. Thanks.

@EricGao888
Copy link
Member

Yes, I think it should be done too, please let me take another look, thanks.

Hello @yangjf2019, may I ask whether you are still working on this issue? We have received feedback from many users that they get blocked by this issue. IMHO we could have a hot-fix for it at the first step, simply increasing the threshold. Then we could move it further and make it configurable for users. Thanks.

Hot fix #11694

@zhuangchong zhuangchong modified the milestones: 3.0.1, 3.0.2 Sep 18, 2022
@TCGOGOGO
Copy link

How about make this configurable? modify code seems tricky

@TheWindIsRising
Copy link

我也遇到了这个错误。请问这个问题是否有任何更新或潜在的解决方案?

你好,请问解决了吗?我也遇到这个问题了,好像是上传时间超过15秒就会自动取消。

image
我的版本是3.1.2
存储位置是MinIO
部署方式是K8s使用heml部署的,请问您是怎么解决的

@TheWindIsRising
Copy link

我也遇到了这个错误。请问这个问题是否有任何更新或潜在的解决方案?

你好,请问解决了吗?我也遇到这个问题了,好像是上传时间超过15秒就会自动取消。

image 我的版本是3.1.2 存储位置是MinIO 部署方式是K8s使用heml部署的,请问您是怎么解决的

我是上传一个超过200多M的时候发现的问题

@TheWindIsRising
Copy link

我的版本是3.1.2请问这个问题前端是否解决?

@TheWindIsRising
Copy link

解决了,需要修改以下4个文件。

/api-server/ui/assets/service.766f4632.js
/api-server/ui/assets/service.766f4632.js.gz

/ui/assets/service.766f4632.js
/ui/assets/service.766f4632.js.gz

并找到以下配置,然后将 15e3 修改为 15e5 或更高版本。baseURL:"/dolphinscheduler",timeout:15e3

@yangjf2019干得好!您想提交 PR 来解决此问题吗?

可否留个联系方式

@TheWindIsRising
Copy link

/api-server/ui/assets/service.766f4632.js

请问这个文件在哪/api-server/ui/assets/service.766f4632.js
由于我是k8s部署的ds,/ui/assets/service.766f4632.js这个文件我在api的pod里面的ui找到了,但是上面那个不知道在哪

@Jacky-YC
Copy link

Jacky-YC commented Apr 22, 2023

image

my version 3.1.5

From the results, more than 15 seconds of requests will be canceled

@dalaopo
Copy link

dalaopo commented Jul 26, 2023

/api-server/ui/assets/service.766f4632.js

请问这个文件在哪/api-server/ui/assets/service.766f4632.js 由于我是k8s部署的ds,/ui/assets/service.766f4632.js这个文件我在api的pod里面的ui找到了,但上面那个不知道在哪

docker没有 ui/assets/下文件,忽略

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment