)]}'
{"specs/ocata/approved/hyper-v-cluster.rst":[{"author":{"_account_id":15130,"name":"Sonu","email":"sonu.sudhakaran@gmail.com","username":"sonu"},"change_message_id":"ddc229bb217c578f9c50ed195dd93720c3369930","unresolved":false,"context_lines":[{"line_number":40,"context_line":"There are two methods for creating and deploying a Hyper-V Cluster, each with"},{"line_number":41,"context_line":"their own advantages and disadvantages:"},{"line_number":42,"context_line":""},{"line_number":43,"context_line":"* Option A. Hyper-V Cluster controlled by a single nova-compute service. This"},{"line_number":44,"context_line":"  means that the nova-compute service will run on a single Hyper-V Node in a"},{"line_number":45,"context_line":"  Cluster and can manipulate WMI objects remotely on all the Cluster Nodes."},{"line_number":46,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a3c09a3_dfb1c802","line":43,"range":{"start_line":43,"start_character":56,"end_line":43,"end_character":63},"updated":"2017-01-17 14:29:34.000000000","message":"nova-compute service itself should be made highly available to make it most of the cluster. Else, if the compute hosting nova service goes down, you make the entire cluster un-available for openstack, which is bad.","commit_id":"98a8bf35d08c78f1fbe96e099fc45117859db0d3"},{"author":{"_account_id":15130,"name":"Sonu","email":"sonu.sudhakaran@gmail.com","username":"sonu"},"change_message_id":"ddc229bb217c578f9c50ed195dd93720c3369930","unresolved":false,"context_lines":[{"line_number":136,"context_line":"healh issues, etc.), moving the instance to another suitable host in the"},{"line_number":137,"context_line":"cluster. This is done automatically by the Microsoft Cluster Service, without"},{"line_number":138,"context_line":"nova\u0027s knowledge. In order to address this issue, there are a few solutions"},{"line_number":139,"context_line":"available:"},{"line_number":140,"context_line":""},{"line_number":141,"context_line":"1. Create a WMI event listener which detects whenever failover occurs, and"},{"line_number":142,"context_line":"notify nova when they occur."}],"source_content_type":"text/x-rst","patch_set":1,"id":"7a3c09a3_bff4fc5a","line":139,"range":{"start_line":139,"start_character":0,"end_line":139,"end_character":9},"updated":"2017-01-17 14:29:34.000000000","message":"Which option is preferred? Honestly, I like the WMI notifying nova of state changes.\n\nWhen a VM has got migrated, I was thinking why the new host compute service won\u0027t detect the new VM? Will this not happen automatically? Do we need a event listener?","commit_id":"98a8bf35d08c78f1fbe96e099fc45117859db0d3"},{"author":{"_account_id":8213,"name":"Claudiu Belu","email":"cbelu@cloudbasesolutions.com","username":"claudiub"},"change_message_id":"3e5f8e121ad4c7cc8b14706dc9ce14e12ebe17c8","unresolved":false,"context_lines":[{"line_number":139,"context_line":"available:"},{"line_number":140,"context_line":""},{"line_number":141,"context_line":"1. Create a WMI event listener which detects whenever failover occurs, and"},{"line_number":142,"context_line":"notify nova when they occur."},{"line_number":143,"context_line":""},{"line_number":144,"context_line":"2. Disable failover migrations entirely, create a WMI event listener which"},{"line_number":145,"context_line":"will detect whenever a failover migration needs to happen, and request nova"}],"source_content_type":"text/x-rst","patch_set":1,"id":"bacf61ea_197017dc","line":142,"updated":"2016-08-03 13:16:13.000000000","message":"they occured*","commit_id":"98a8bf35d08c78f1fbe96e099fc45117859db0d3"}]}