Description of problem: During sync, if you run into problems with bad repo metadata, the deployment continues as if nothing is wrong and no errors are reported to the user. Version-Release number of selected component (if applicable): Nightly How reproducible: Always. Steps to Reproduce: 1. Mirror the Red Hat CDN 2. Mess with the checksums in a repomd.xml file 3. Point your RHCI app to the mirrored CDN (via the Subscriptions page) 4. Perform a deployment Actual results: Deployment continues and no errors are reported to the user until the provisioning fails. The dynflow tasks are reported as "success" but the sync status page says "New packages: 9 (921 KB). Failed to download 9 packages." Expected results: The dynflow task should error.
I can confirm that this bug exist. I was doing a RHEV deployment yesterday and my deployment got stuck at 91.2%. In the Foreman deployment.log, all I have is "Rhev::WaitForDataCenter get_status method". When I go to "Content -> Sync Status" in the Satellite interface, I can see that the repo for RHEV did not sync correctly and show an error message : "Yum Metadata: [Errno 2] No such file or directory: u'/var/lib/pulp/working/repos/Default_Organization-Red_Hat_Enterprise_Virtualization-Red_Hat_Enterprise_Virtualization_Manager_3_5_RPMs_x86_64_6Server/importers/yum_importer/tmpdC20BY/44065fd3e22b2eebbfaf64dc6c970bf11cd1d3cc-filelists.xml.gz'" All the other repos were synced correctly. To solve the issue, I just started the synchronisation back again and the deployment completed successfully.
This was fixed when we moved to Sat 6.1.7
Verified on QCI-1.0-RHEL-7-20160411.t.0.
QCI has sunset. Closing the BZ's.