The "charms tests" section in any test results (http://reports.vapour.ws/releases/5121) show we have comprehensive testing across substrates. We know know joyent is problematic at this time.
Given that this is likely a cloud-int or cloud-image problem, we are not likely to see this issue in images that we have 100% control like AWS, Azure, MAAS, vsphere, lxd. The three clouds that alter our images (and delay releases) are Joyent, Rackspace, and GCE.
I feel pretty powerless in this matter. I know the test will pass if I could force apt-get update. I presume juju tried and it was ignore. I could modify the charm or the bundle to trigger apt-get update. This act would undermine the landscape (not juju) in this test. Bad images are a cloud-health issue and I prefer to be catching this class on issue in cloud-heatlh checks, not juju revision tests.
The "charms tests" section in any test results (http:// reports. vapour. ws/releases/ 5121) show we have comprehensive testing across substrates. We know know joyent is problematic at this time.
Reviewing http:// reports. vapour. ws/releases/ issue/5553b2797 49a56508ba88b45 for specific substrates. We can see that rackspace was pretty busted Feb 1-8. GCE which, I thought has recent failures, not so (Mary 2016).
Given that this is likely a cloud-int or cloud-image problem, we are not likely to see this issue in images that we have 100% control like AWS, Azure, MAAS, vsphere, lxd. The three clouds that alter our images (and delay releases) are Joyent, Rackspace, and GCE.
I feel pretty powerless in this matter. I know the test will pass if I could force apt-get update. I presume juju tried and it was ignore. I could modify the charm or the bundle to trigger apt-get update. This act would undermine the landscape (not juju) in this test. Bad images are a cloud-health issue and I prefer to be catching this class on issue in cloud-heatlh checks, not juju revision tests.