Description of problem: During an upgrade of RHEV-3.6.10 (upgraded since RHEV-3.3) and additionally migrating from out-of-band cluster for the manager to self-hosted, we ran into the issue that the manager was not able to add the hosted-engine host. Version-Release number of selected component (if applicable): * Running RHEV-3.6.10 with "rhevm" as management bridge * Upgrading to RHV-4.0 by installing ovirt-hosted-engine-setup and bringing up a new engine, applying the backup from RHEV-3.6.10 How reproducible: Always Steps to Reproduce: 1. Install RHEV-manager with "rhevm" as management bridge (has been default until 3.5/3.6??) 2. Create backup of database 3. Install RHV-4 host with hosted-engine (a separated BZ will be filed because ovirt-hosted-engine-setup pulls in RHV-4.1 appliance during setup) 4. Restore backup from RHEV-3.6.10 on RHV-4.0 engine for upgrade 5. run "engine-setup" 6. shutdown engine as directed when engine-setup finished. 7. find ovirt-hosted-engine-setup starting the engine with "ovirtmgmt" as management interface whereas the hosts use "rhevm". Actual results: ovirt-hosted-engine-setup creates unconditionally "ovirtmgmt" as bridge for the management interface. Expected results: ovirt-hosted-engine-setup asks if the installation is an upgrade and if so asks for the name of the currently used management bridge. Also the question about starting "engine-setup" could then be defaulting to "no". Additional info:
There is a workaround for that described in https://access.redhat.com/solutions/2292861 *** This bug has been marked as a duplicate of bug 1369690 ***