Skip to content

Conversation

@SamWhited
Copy link
Contributor

#2419 and
#2407
attempted to seperate out empty parent and internal for
macvlan and ipvlan networks

However it didnt pass the integration tests in moby
moby/moby#40596 and exposed some
more plumbing that needed to be done to make sure
we separate the two things

If the -o parent is empty we create a dummylink
and if internal is set we dont add a default gateway
and make sure north-south communication cannot take place
(only east-west / container-container can)

Signed-off-by: Arko Dasgupta arko.dasgupta@docker.com

#2419 and
#2407
attempted to seperate out empty parent and internal for
macvlan and ipvlan networks

However it didnt pass the integration tests in moby
moby/moby#40596 and exposed some
more plumbing that needed to be done to make sure
we separate the two things

If the -o parent is empty we create a dummylink
and if internal is set we dont add a default gateway
and make sure north-south communication cannot take place
(only east-west / container-container can)

Signed-off-by: Arko Dasgupta <arko.dasgupta@docker.com>
@SamWhited
Copy link
Contributor Author

Cherry picking this over as I think it's needed to make tests pass on the moby/19.03 branch in moby/moby#40617

@trapier
Copy link

trapier commented Mar 11, 2020

Commenting to indicate for the record that this is a backport of #2523

@arkodg
Copy link
Contributor

arkodg commented Mar 11, 2020

@SamWhited I don't think we want to add these changes to 19.03
the DNS vendoring is the only relevant cherry-pick for 19.03 IMHO

the above fixes should be available in 20.03/20.04 which is soon :)

@SamWhited
Copy link
Contributor Author

Someone had mentioned they thought this is what was causing some of the test failures we were seeing, but I didn't look into it too far. I've been bouncing back and forth between two issues today, so let me see if I can confirm or deny that and get back to you.

@trapier
Copy link

trapier commented Mar 11, 2020

@arkodg confusion is probably my fault. I misread the log on bump_19.03 and thought #2419 and #2407 had been backported. If that's not the case then this PR is indeed unnecessary.

For full context our interest is to see #2525 in moby/moby 19.03.

@SamWhited
Copy link
Contributor Author

Ah my mistake, I mixed up the commits again too. Also, the test failures I was thinking of seem to have magically been fixed and I have no idea why so :nothingtodohere:.

@SamWhited SamWhited closed this Mar 11, 2020
@SamWhited SamWhited deleted the bump_19.03 branch March 11, 2020 20:05
@arkodg
Copy link
Contributor

arkodg commented Mar 11, 2020

ah so we don't need anything else in libnetwork:bump_19.03 and need the new refpoint to be vendored into moby:19.03
Ideally this PR moby/moby#40617 can be updated with the new refpoint to include #2525

@SamWhited
Copy link
Contributor Author

That PR appears to be up to date with the latest commit from bump_19.03

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

Successfully merging this pull request may close these issues.

3 participants