We should limit the thread pool size for host upgrade checks
Verified with rhevm-4.1.1.2-0.1.el7.noarch. Have 6 hosts in my env, all are 'UP'. restart ovirt-engine service. After 15 minutes the system invokes the host upgrade check on the hosts. the 1st 5 host are checked async, the 6th host is checked only after the first host's check was done, and uses the same thread as the fisrt one had.