Hide Forgot
Created attachment 551917 [details] Screenshot of the final error message Description of problem: Deployment of an image fails with confusing error message, if there is no catalog created. Version-Release number of selected component (if applicable): 0.8.0-2 How reproducible: Bug is easy to reproduce. Preliminary steps: 1. Install aeolus-all-0.8.0-2.el6 on RHEL 6.2 2. Turn off SELinux Enforcing mode and update the firewall. 3. Run /usr/sbin/aeolus-configure -p ec2,mock 4. Log into the Web UI. 5. Setup Amazon EC2 credentials. Steps to Reproduce: 1. Click on card "Administer" in the Web UI 2. Click on "Environments" 3. Click on "New Image" in "default" pool family. 4. Write arbitrary string into the "Name" entry field. 5. Add any valid a template file by "Choose a template file" entry (e.g. example template from the online documentation). 6. Click on "Continue" 7. Click on "Save and Continue" 8. Click on "Save Template" 9. Click on "Save" Actual results: Getting following two confusing messages (see attachment): * "Deployable added" * "Deployable was not created: No route matches {:action=>"show", :controller=>"catalogs"}" Expected results: I would expect an error message right after the step 3 that prevents user to proceed up to step 9. Additional info: There is no catalog available in the fresh installation of Aeolus and new image can be no way added in such circumstances. That is the reason why user should never get beyond step 3 if there is no catalog created. Suggested fixes: * Add a check for the number of catalogs available and do not let user to go beyond step 3 in case if number of catalogs is lower than 1. * Alternatively, the configure script can preventively create the "default" catalog
adding to sprint tracker
We will address this by ensuring that aeolus-configure creates a default catalog. *** This bug has been marked as a duplicate of bug 781474 ***
Bug#781474 discusses having a catalog created by default by aeolus-configure. As I understand, this works around this problem, but doesn't address it. If the user chooses to delete all catalogs, they are open to the this problem. I believe this bug should prevent the user from entering a new image if no catalogs exist. I propose decoupling this as a DUPLICATE bug
I think I mentioned the same concern in the initial conversation where we decided to dupe it. I'm not opposed to tracking this separately as a distinct issue.
(In reply to comment #4) > I think I mentioned the same concern in the initial conversation where we > decided to dupe it. I'm not opposed to tracking this separately as a distinct > issue. Thanks Mike ... I'm opening this backup. Sounds like the urgency of this problem will be mitigated by creating a default catalog upon running aeolus-configure (bug#781474). However, the user is still exposed to shooting themselves in the foot and removing all catalogs. With no catalog created, the user is allowed to create an image. During image creation, if the user accepts the default-selected option ([X] Automatically make "client2" deployable), they are able to ... 1. Successfully create the image 2. Fail to create the deployable I think this bug should be repurposed to disable the option "Automatically make "client2" deployable." if there are no catalogs defined. Thoughts/concerns?
*** Bug 785010 has been marked as a duplicate of this bug. ***
fix has been posted on aeolus-devel.org https://fedorahosted.org/pipermail/aeolus-devel/2012-February/008536.html
commit 8a80b0f785e0c5e4c510e875cf8443b62d5671a0 solved this issue as it redesigns image creation workflow, so the bug is not present any more. Previously posted fix is obsolete now.
8a80b0f in aeolus-conductor-0.8.0-27
Verified this bug: The fix has removed the "Automatically make "name" deployable" feature when you upload(upload,from URL) a component outline from UI even if the catalog is present. So now the issue has been resolved.if no catalogs are present User is displayed with this "No catalog exists! Please create one."message during deployable creation. info: [root@nec-em25 html]# rpm -qa | grep aeolus rubygem-aeolus-image-0.3.0-10.el6.noarch aeolus-conductor-daemons-0.8.0-36.el6.noarch aeolus-conductor-doc-0.8.0-36.el6.noarch aeolus-configure-2.5.0-15.el6.noarch rubygem-aeolus-cli-0.3.0-10.el6.noarch aeolus-all-0.8.0-36.el6.noarch aeolus-conductor-0.8.0-36.el6.noarch
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. http://rhn.redhat.com/errata/RHEA-2012-0583.html