Please use code tags, as required by rules.
After 100 posts on these forums this should be endorsed while posting.
As for your question, if a resource group is in Online state, the package will fail over upon host failure, if quorum is met and the resource group has remaining node(s) in its configuration.
Can you post output in code tags of :
clrg show cognos10-rg
Unmanaged state will do nothing, this means that the package is configured (probably, i can only guess without further inspection) but is not managed by cluster.
So, host failures or anything will not effect those in a
cluster sense.
Best is, after migration to clean the old resources and resource groups from cluster, and old devices files from operating system.
Also, good practice is to trying to failover a package before going live
Otherwise inspect the system logs, and see if those packages failovered in the past, giving you further clues.
Hope that helps.
Regards
Peasant.