A BZ came in that requested we open up read permission on our AMIs
Sayan wrote a script to do this:
https://lists.fedoraproject.org/archives/list/cloud@lists.fedoraproject.org/message/GOXAHJLWG3EMD47Y5IW4CMK4H4PHSIRA/
Where are we on getting this done?
@maxamillion has reported for several weeks that this in "in-flight" with @ausil mentioning something with regards to credentials. @ausil can you give us an update?
Metadata Update from @dustymabe: - Issue assigned to sayanchowdhury
My PR to libcloud to modify snapshot attribute was merged. So I will fix it in fedimg too and do a release.
yay! - can you modify existing images with the tool?
@sayanchowdhury - can we get an update?
The patch is included in the latest release "2.0.0rc2". But, we need to package the latest release before we can go ahead.
I emailed @dbruno who is the maintainer of the python-libcloud package. He will be working on the releasing the updated package today itself.
I've pushed the update to test [0], feel free to test and give your feedback.
[0] - https://bodhi.fedoraproject.org/updates/FEDORA-2017-435fedd281
@dbruno can you build the package for el7 also? I did a scratch build and tested the package. The patch seems to be working.
I have created a PR for this issue in fedimg: https://github.com/fedora-infra/fedimg/pull/55/
Metadata Update from @dustymabe: - Issue tagged with: infra
@pnemade built libcloud for el7. I have cut out a new release for fedimg too. If everything looks good, the fix should be out in 1-2 days (after FBR).
The fix has been deployed to production. I will keep this issue open for the weekend just to see everything turns out fine.
thanks sayan!
Good news - we can now create snapshots and new AMIs from out AMIs - bad news - non us-east-1 images are broken: https://pagure.io/atomic-wg/issue/286
closing this.
Metadata Update from @dustymabe: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Seems like few of the snapshots were not getting the proper permission. Fix has been deployed. Should be fine now.
Log in to comment on this ticket.