-
Notifications
You must be signed in to change notification settings - Fork 880
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
Overlay driver appears to be leaking veth devices #984
Comments
I suspect this is in the |
I see this issue too and had to reboot all Swarm hosts to get rid of the 300+ veth devices per host that accumulated and were driving the monitoring system crazy. Now they are building up again. |
You can run this as a quick fix:
it cleans all downed veth interfaces on your docker host |
@thomasem
|
@thomasem also, can you paste the |
Awesome. Thanks @aboch! |
Using Docker Swarm v1.1.0, Docker 1.10.1
From https://github.com/getcarina/feedback/issues/40:
network to see this problem.
Steps to recreate:
Start with a fresh one-segment cluster and get a baseline for already
existing devices on the segment:
Create a new overlay network:
Spin up two new containers on this network:
Check the segment's network namespace to see the newly created devices,
docker_gwbridge
,vethb5edef0
, andveth79df616
:Delete one of the containers that was created and attached to the
test-veth
network:Check the segment's network namespace again, and notice that the list has
the same number of veth devices as before, where it should be one short, and
observe that
veth79df616
is gone, replaced byveth1d6abe9
.Delete the last container that was spun up on the
test-veth
overlaynetwork:
Check the segment's network namespace again and observe the same behavior,
except this time
vethb5edef0
is replaced byveth0ad699c
:Double check that all containers on that network were deleted:
The text was updated successfully, but these errors were encountered: