New commit detected on ManageIQ/manageiq/gaprindashvili: https://github.com/ManageIQ/manageiq/commit/6501817774fc8c83bda0047cc90ac39991090d85 commit 6501817774fc8c83bda0047cc90ac39991090d85 Author: Nick Carboni <ncarboni> AuthorDate: Tue Apr 3 10:25:46 2018 -0400 Commit: Nick Carboni <ncarboni> CommitDate: Tue Apr 3 10:25:46 2018 -0400 Merge pull request #17222 from robbmanes/properly_validate_proxy_urls Resolve string handling for "https://" or "http://" in update_rhsm_conf (cherry picked from commit b44f6b17977f8f8e317360ecde03ea92e35b32ba) https://bugzilla.redhat.com/show_bug.cgi?id=1566562 app/models/registration_system.rb | 5 +- spec/models/registration_system_spec.rb | 35 +- 2 files changed, 24 insertions(+), 16 deletions(-)
So this seems to be partially fixed. The proxy info gets updated in the rhsm.conf correctly now but registration still has some issues. From what I am seeing registration does seem to be completed checking evm.log but the last message in the UI gets stuck at 'Enabling rhel-server-rhscl-7-rpms' and never moves to 'Registered'. Also this only seems to happen when you have a proxy configured and no access on 443 and 80 ports.
Looking through the log, everything appears to be working as expected with the proxy. The performance of the subscription-manager commands seems to be another issue (taking 3+ minutes to "subscription-manager register") and may warrant another bug. It seems this is causing our worker to hit the 10 minute timeout and get killed before it can finish all of the steps in the registration process.
Well in the logs it seems like registration is complete, but the UI last message never gets to 'Registered'. Also there is no websocket notification with 'completed successfully'. I feel this is still a problem just from the fact that it gives confusion to the user, there is no way to tell that your appliance is completely registered without digging through logs. It just feels like its stuck in limbo. I am happy to move this to verified if we are sure registration has finished, but if the worker is timing out surely its not, in which case something else could be missing. If someone has a slow proxy/network for instance it could time out before attaching all repos, correct?
We only emit a notification if registration fails and it seems we are missing a last message update for "completed successfully". Registration is queued to happen at least once a day by a schedule, so if it gets killed in the middle (due to extremely slow connection) it will pick up where it left off last time and the status will get updated as well.
Does that mean it should eventually get the "registered" message in the UI if it continues where it left off each day or will it always timeout for w/e reason? I am just concerned there might be some underlined issue we are not seeing, but I have tried checking for updates and it does find them so it seems like everything is working. I will move this to verified and raise a new bug for the notifications/last message stuff. Thanks for the help Brandon.
Yes, I would expect that subsequent calls to subscription-manager wouldn't take as long if it is already registered, so that should take several minutes off of the process. The underlying issue is that the subscription-manager commands are taking a very long time to complete.
New commit detected on ManageIQ/manageiq/master: https://github.com/ManageIQ/manageiq/commit/39828bca80afd43f35245e496d5b144e18597b3c commit 39828bca80afd43f35245e496d5b144e18597b3c Author: Brandon Dunne <brandondunne> AuthorDate: Tue Apr 24 12:04:12 2018 -0400 Commit: Brandon Dunne <brandondunne> CommitDate: Tue Apr 24 12:04:12 2018 -0400 Add logging and update upgrade_message for successful registration It seems hard to tell whether the entire process of registering, attaching products and enabling repos has finished. https://bugzilla.redhat.com/show_bug.cgi?id=1566562 app/models/miq_server/update_management.rb | 2 + 1 file changed, 2 insertions(+)
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2018:1328