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

Fix wrong DNS name in certificates issued to HA controllers. #7748

Closed
wants to merge 1 commit into from
Closed

Fix wrong DNS name in certificates issued to HA controllers. #7748

wants to merge 1 commit into from

Conversation

Pekkari
Copy link

@Pekkari Pekkari commented Aug 15, 2017

This PR fixes the DNS name issued in server certificate that prevents
a juju agent to connect properly to mongodb database when on an HA
environment.

lp:1710886

Signed-off-by: José Pekkarinen jose.pekkarinen@canonical.com

Signed-off-by: José Pekkarinen <jose.pekkarinen@canonical.com>
Copy link
Member

@jameinel jameinel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you have a specific test case (be it a feature test, or some other check) that fails without this change but succeeds with this change?
I'm not sure where the string 'local' comes from, but I believe the other strings are never "real" DNS names, but just names that get put into the certificates/advertised. (we allow juju controllers to advertise their name as simply 'juju-apiserver', since all the user-bootstrapped apiserver's won't have a valid DNS entry anyway.
I did try grepping around to see if there was some other place where we were naming something 'local', but I don't see it.

That particular line seems to go all the way back to:
https://bugs.launchpad.net/juju-core/+bug/1434680
and
https://github.com/juju/juju/pull/1906/files

Looking at that diff, it does appear that other places use the full 'localhost'
https://github.com/juju/juju/pull/1906/files#diff-45200f2105c0995c08150ba75443b6fbR108

It appears that maybe this code only triggers during 'upgrade' and not during initial bootstrap, etc?

It does look correct to me, but ideally there would be some sort of test associated with it, so we don't regress it in the future.

@jameinel
Copy link
Member

Also, I should note that we're unlikely to release another 2.1 at this stage. I think we would be likely to release a 2.2 series, so we could target the fix there. If there is a strong reason why you need the fix to be in 2.1 we can certainly discuss it, but ATM there are no plans to release a 2.1.4.

@wallyworld
Copy link
Member

The fix certainly looks like it solves a legitimate problem. Give that it's on a branch which has been superseded by a later release, I think it would be acceptable for an assurance that manual testing has been done to confirm the breakage before the fix, and correct behaviour after the fix.

@Pekkari
Copy link
Author

Pekkari commented Aug 15, 2017

@howbazaar
Copy link
Contributor

Sorry, not doing another 2.1 release. following #7755.

@howbazaar howbazaar closed this Sep 6, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
4 participants