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

rd0 ip routing prioritisation appears to result in slow external network downloads #1070

Closed
nathanpiper opened this issue Dec 6, 2021 · 4 comments · Fixed by #1094
Closed
Assignees
Labels
kind/bug Something isn't working
Milestone

Comments

@nathanpiper
Copy link

nathanpiper commented Dec 6, 2021

Rancher Desktop Version

main 4c29159

Rancher Desktop K8s Version

1.21.7

What operating system are you using?

macOS

Operating System / Build Version

12.0.1

What CPU architecture are you using?

x64

Windows User Only

No response

Actual Behavior

Experiencing really slow external network downloads on the VM. I first notice when image pulls appeared to be taking an extremely long time to complete.

Steps to Reproduce

  1. Shell into VM instance and download a test file using wget:

wget --output-document=/dev/null https://ftp.zetta.net.au/pub/testfiles/testfile.300Mb

Experience slow performance

  1. Delete prioritised ip route:
    ip route del default via 192.168.205.1 dev rd0 metric 201

Experience expected (faster) performance

Result

wget --output-document=/dev/null https://ftp.zetta.net.au/pub/testfiles/testfile.300 Mb Connecting to ftp.zetta.net.au (119.252.91.137:443) saving to '/dev/null' null 0% | | 992k 0:53:10 ETA

Expected Behavior

wget --output-document=/dev/null https://ftp.zetta.net.au/pub/testfiles/testfile.300 Mb Connecting to ftp.zetta.net.au (119.252.91.137:443) saving to '/dev/null' null 3% |*** | 10.1M 0:01:54 ETA^

Additional Information

Appears to be introduced by the rd0 prioritisation scripting on this PR: #974

@nathanpiper nathanpiper added the kind/bug Something isn't working label Dec 6, 2021
@nathanpiper nathanpiper changed the title rd0 ip routing priority appears to result in slow external network downloads rd0 ip routing prioritisation appears to result in slow external network downloads Dec 6, 2021
@nathanpiper
Copy link
Author

I would love to help out here, but I'm a bit out of my depth on the networking configuration.

@mattfarina
Copy link
Contributor

@jandubois can you please take a look at this.

@jandubois
Copy link
Member

@jandubois can you please take a look at this.

Yes, already on my list to repro today.

If I can confirm, I will try to implement "point-to-point" mode for vde_vmnet. It eliminates the vde_switch, but means only a single VM will be able to connect to the vde_vmnet daemon (which is fine for Rancher Desktop because we only run a single VM so far). I don't know if this would fix things though, so just something to try.

@jandubois
Copy link
Member

I've reproduced the issue, and have filed 2 PRs that will improve the situation somewhat:

virtualsquare/vde-2#35
lima-vm/vde_vmnet#31

The first one includes timings from my tests; the fixes improved download times of a 4GB ISO from 75min to 90s, which sounds awesome until you notice that it only takes 13s over slirp (and 3s directly on the host).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
kind/bug Something isn't working
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants