bug 1702211 was opened around the same time frame, and we landed a fix for it in Pike. I wonder if that had something to do with this report, too. It certainly was a transient issue based on SQL. We fixed it by migrating to a more precise date format and making keystone responsible for marshaling the date to something it expects instead of letting SQL handle things.
bug 1702211 was opened around the same time frame, and we landed a fix for it in Pike. I wonder if that had something to do with this report, too. It certainly was a transient issue based on SQL. We fixed it by migrating to a more precise date format and making keystone responsible for marshaling the date to something it expects instead of letting SQL handle things.