Issue

After restarting the engine service can start VMs only from one cluster.
Start or migrate VMs in other cluster fails with error: no available running hosts with cluster displayed network.

Looking into engine.log:

2014-05-02 16:48:30,552 INFO  [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp-/127.0.0.1:8702-1) [70ec00e6] 
Candidate host rhev1 (d0eb8853-b32a-41a6-b814-93d4e287662b) was filtered out by VAR__FILTERTYPE__INTERNAL filter Network
2014-05-02 16:48:30,553 INFO  [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp-/127.0.0.1:8702-1) [70ec00e6] 
Candidate host rhev2 (d3e62845-71c4-44b3-b6c1-01be6fa907b2) was filtered out by VAR__FILTERTYPE__INTERNAL filter Network
2014-05-02 16:48:30,553 WARN  [org.ovirt.engine.core.bll.RunVmCommand] (ajp-/127.0.0.1:8702-1) [70ec00e6] 
CanDoAction of action RunVm failed. Reasons:VAR__ACTION__RUN,VAR__TYPE__VM,ACTION_TYPE_FAILED_MISSING_DISPLAY_NETWORK,ACTION_TYPE_FAILED_MISSING_DISPLAY_NETWORK,SCHEDULING_ALL_HOSTS_FILTERED_OUT,VAR__FILTERTYPE__INTERNAL,$hostName rhev1, ...