)]}'
{"nova/conf/scheduler.py":[{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"ea695540b2e941fb263c38a756a52945fcaa86d1","unresolved":true,"context_lines":[{"line_number":63,"context_line":"  periodic tasks."},{"line_number":64,"context_line":"\"\"\"),"},{"line_number":65,"context_line":"    cfg.IntOpt(\"max_placement_results\","},{"line_number":66,"context_line":"        default\u003d1000,"},{"line_number":67,"context_line":"        min\u003d1,"},{"line_number":68,"context_line":"        help\u003d\"\"\""},{"line_number":69,"context_line":"The maximum number of placement results to request."}],"source_content_type":"text/x-python","patch_set":2,"id":"be0779d7_8e26bf12","line":66,"range":{"start_line":66,"start_character":16,"end_line":66,"end_character":20},"updated":"2021-02-16 21:03:39.000000000","message":"I\u0027m sure there\u0027s a reason it\u0027s this high and not something lower that we are asking larger environments to raise but wow that\u0027s huge. Why would anyone ever need 1000 candidates during scheduling?","commit_id":"737cae76e4c7c08e94a5bfe3288641e03e9e7023"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"b54d69c201a167a14f652ab3475f5d07cf612437","unresolved":true,"context_lines":[{"line_number":63,"context_line":"  periodic tasks."},{"line_number":64,"context_line":"\"\"\"),"},{"line_number":65,"context_line":"    cfg.IntOpt(\"max_placement_results\","},{"line_number":66,"context_line":"        default\u003d1000,"},{"line_number":67,"context_line":"        min\u003d1,"},{"line_number":68,"context_line":"        help\u003d\"\"\""},{"line_number":69,"context_line":"The maximum number of placement results to request."}],"source_content_type":"text/x-python","patch_set":2,"id":"6d9bc786_c17a8c80","line":66,"range":{"start_line":66,"start_character":16,"end_line":66,"end_character":20},"in_reply_to":"be0779d7_8e26bf12","updated":"2021-03-07 10:12:53.000000000","message":"As we still have a split decision between placement and nova scheduler about the target host we have the edge case where the placement a_c query could return every compute host but the nova schedule (e.g. the NumaTopologyFilter or the PCIFilter) would filter out all but a couple of compute hosts from that result. So keep this low means potential NoValidHost results in an env where there are available resources to fulfill the boot request. \n\nAnother reason to keep this number high is that placement returns more than one candidates per host (e.g. when there is multiple devices with bandwidth resource configured) then nova today simply use one of those candidates to filter the host. So asking for 100 candidate might mean we get info about 25 hosts (assuming 4 PFs with bandwidth on each host).","commit_id":"737cae76e4c7c08e94a5bfe3288641e03e9e7023"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"ea695540b2e941fb263c38a756a52945fcaa86d1","unresolved":false,"context_lines":[{"line_number":253,"context_line":"Instances performing IO includes those in the following states: build, resize,"},{"line_number":254,"context_line":"snapshot, migrate, rescue, unshelve."},{"line_number":255,"context_line":""},{"line_number":256,"context_line":"Note that this setting only affects scheduling if the ``IoOpsFilter`` filter is"},{"line_number":257,"context_line":"enabled."},{"line_number":258,"context_line":""},{"line_number":259,"context_line":"Possible values:"}],"source_content_type":"text/x-python","patch_set":2,"id":"ca9b5fc7_098fba73","line":256,"range":{"start_line":256,"start_character":56,"end_line":256,"end_character":67},"updated":"2021-02-16 21:03:39.000000000","message":"TIL","commit_id":"737cae76e4c7c08e94a5bfe3288641e03e9e7023"},{"author":{"_account_id":10135,"name":"Lee Yarwood","display_name":"Lee Yarwood","email":"lyarwood@redhat.com","username":"lyarwood"},"change_message_id":"ea695540b2e941fb263c38a756a52945fcaa86d1","unresolved":true,"context_lines":[{"line_number":604,"context_line":"  which the instance is currently running."},{"line_number":605,"context_line":""},{"line_number":606,"context_line":"Related options:"},{"line_number":607,"context_line":""},{"line_number":608,"context_line":"* ``[filter_scheduler] weight_classes``"},{"line_number":609,"context_line":"\"\"\"),"},{"line_number":610,"context_line":"    cfg.BoolOpt("}],"source_content_type":"text/x-python","patch_set":2,"id":"581aba2a_aae8fca4","line":607,"updated":"2021-02-16 21:03:39.000000000","message":"Can we reference the compute:servers:resize:cross_cell policy here or the docs under https://docs.openstack.org/nova/latest/admin/configuration/cross-cell-resize.html somehow?","commit_id":"737cae76e4c7c08e94a5bfe3288641e03e9e7023"}]}