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] Error for using graphscope in “hosts”mode. #2887

Open
Pomelochen opened this issue Jun 14, 2023 · 12 comments
Open

[BUG] Error for using graphscope in “hosts”mode. #2887

Pomelochen opened this issue Jun 14, 2023 · 12 comments

Comments

@Pomelochen
Copy link

Describe the bug
A clear and concise description of what the bug is.
When I try to create a session in local mode, I meet an error, here is the debug log :

Screenshots
IMG_4440

Environment (please complete the following information):

  • GraphScope version: [0.21]
  • OS: [Linux rehhat 7.5 and centos7]
    -python:3.10

Additional context
We don’t meet the same bug in k8s mode.

@welcome
Copy link

welcome bot commented Jun 14, 2023

Thanks for opening your first issue here! Be sure to follow the issue template! And a maintainer will get back to you shortly!
Please feel free to contact us on DingTalk, WeChat account(graphscope) or Slack. We are happy to answer your questions responsively.

@siyuan0322
Copy link
Collaborator

As could be seen in the pic, 0.0.0.0:2379: bind: address already in use
I surmise some old etcd process is running on your machine?

@Pomelochen
Copy link
Author

As could be seen in the pic, 0.0.0.0:2379: bind: address already in use
I surmise some old etcd process is running on your machine?

The reason you said I have considered. I have checked the pod 2379, it is free. And I run the programmer in a docker container. So the address is free.

@siyuan0322
Copy link
Collaborator

But I couldn't see any other reasons that could cause the address already in use error 😂
Try the following:

  1. nc -vl 2379 to ensure the port could be listened, then cancel the nc by Ctrl-C
  2. kill previous etcd process pkill etcd
  3. Start graphscope again.

Hope you could try the above ways and share the results.

@Pomelochen
Copy link
Author

But I couldn't see any other reasons that could cause the address already in use error 😂 Try the following:

  1. nc -vl 2379 to ensure the port could be listened, then cancel the nc by Ctrl-C
  2. kill previous etcd process pkill etcd
  3. Start graphscope again.

Hope you could try the above ways and share the results.

I test nc -vlp 2379, ensure the port could be listened. Then conduct pill etcd , start the gs again. The same bug occurred.

@sighingnow
Copy link
Collaborator

@Pomelochen Try:

graphscope.set_option(etcd_listening_client_port=2479)
graphscope.set_option(etcd_listening_peer_port=2480)

Before creating the session. See also #1903 and #1910.

@Pomelochen
Copy link
Author

@Pomelochen Try:

graphscope.set_option(etcd_listening_client_port=2479)
graphscope.set_option(etcd_listening_peer_port=2480)

Before creating the session. See also #1903 and #1910.

Unfortunately, I try to change the port as you said. And other ports, the same bug address already in use.

@sighingnow
Copy link
Collaborator

And other ports, the same bug address already in use.

Can you please paste the error message you have?

@sighingnow
Copy link
Collaborator

And other ports, the same bug address already in use.

Can you please paste the error message you have?

I have made some changes to reduce the dependency of analytical engine on etcd in #2889, which is expected to be available in tomorrow's nightly build. By then, you could upgrade your graphscope installation by

pip3 install -U grpahscope --pre -i https://pypi.org/simple

and try again (the nightly build will be available tomorrow, not today).

sighingnow added a commit that referenced this issue Jun 14, 2023
…l sessions (#2889)

Mitigate #2887.

Signed-off-by: Tao He <linzhu.ht@alibaba-inc.com>
@Pomelochen
Copy link
Author

And other ports, the same bug address already in use.

Can you please paste the error message you have?

I have made some changes to reduce the dependency of analytical engine on etcd in #2889, which is expected to be available in tomorrow's nightly build. By then, you could upgrade your graphscope installation by

pip3 install -U grpahscope --pre -i https://pypi.org/simple

and try again (the nightly build will be available tomorrow, not today).

Ok, I will try it. If possible. You can try to make a image with ubuntu20.04, and pip graphscope. And run the image on a machine with os readhat7.5,use graphscope on hosts. You may have the same bug. I have try several machine, meet the same bug.

@sighingnow
Copy link
Collaborator

sighingnow commented Jun 15, 2023

Ok, I will try it.

Nightly CI failed last night, please wait for another day.

Rerunning, expected to be available in an hour.

It's available now.

@Pomelochen
Copy link
Author

And other ports, the same bug address already in use.

Can you please paste the error message you have?

I have made some changes to reduce the dependency of analytical engine on etcd in #2889, which is expected to be available in tomorrow's nightly build. By then, you could upgrade your graphscope installation by

pip3 install -U grpahscope --pre -i https://pypi.org/simple

and try again (the nightly build will be available tomorrow, not today).

When I try to install the new graphscope, I meet another bug
IMG_4470
Graphscope version 0.22.0
Host mode

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants