-
Notifications
You must be signed in to change notification settings - Fork 2.9k
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
What is the real performance of flannel #738
Comments
I think your test is fine. vxlan can do many GBPS so you will only see the performance drop compared to no encapsulation if you have a fast enough network. |
Just to confirm @tomdee , I performed this benchmark: Test Results:
So 23.4GBytes Total data transfered in 30 secs with average 6.68Gbit/sec bandwidth.
So 11.1GBytes Total data transfered in 30 secs with average 3.17Gbit/sec bandwidth. |
Hi all
We are deploying openshift platform and I have some simple investigation on flannel.
Here is our environment:
AWS VPC cn-north-1
CentOS 7.3 with enhanced networking enabled
kernel 3.10.0-514.21.1.el7.x86_64
flannel 0.7.0 with vxlan backend
Below is my procedures:
start iperf3 on host01
iperf -s
iperf test with host ip on another host
iperf -c hostip -t 60
iperf test with overlay ip on another host
iperf -c overlayip -t 60
From our test results I have seen 99.5% performance compared to native host network.
I have done this test for both c4 and m4 instances types, and instances in different subnets and available zones.
However, I searched google and most people say the performance is 50% approximately.
Did I do something wrong in this test ? Or the flannel auto leverage the aws overlay network infrustructure that accelerating the vxlan ?
The text was updated successfully, but these errors were encountered: