)]}'
{"nova/compute/resource_tracker.py":[{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"56663f389afcf7b0233f0bd7d9fc82d338da1318","unresolved":false,"context_lines":[{"line_number":982,"context_line":"        # set some initial values, reserve room for host/hypervisor:"},{"line_number":983,"context_line":"        cn.local_gb_used \u003d CONF.reserved_host_disk_mb / 1024"},{"line_number":984,"context_line":"        cn.memory_mb_used \u003d CONF.reserved_host_memory_mb"},{"line_number":985,"context_line":"        cn.vcpus_used \u003d CONF.reserved_host_cpus"},{"line_number":986,"context_line":"        cn.free_ram_mb \u003d (cn.memory_mb - cn.memory_mb_used)"},{"line_number":987,"context_line":"        cn.free_disk_gb \u003d (cn.local_gb - cn.local_gb_used)"},{"line_number":988,"context_line":"        cn.current_workload \u003d 0"}],"source_content_type":"text/x-python","patch_set":1,"id":"5ff73747_c61bb4c7","line":985,"updated":"2017-04-27 19:30:11.000000000","message":"Note https://review.openstack.org/#/c/457782/2/nova/compute/resource_tracker.py@104 - you might want to rebase on top of that change.","commit_id":"2492ad97b86a4e5a13afcd2fe0f14f635664b69e"}],"nova/conf/compute.py":[{"author":{"_account_id":7,"name":"Jay Pipes","email":"jaypipes@gmail.com","username":"jaypipes"},"change_message_id":"c78215ea5010e05615932ff287723216f45f8041","unresolved":false,"context_lines":[{"line_number":399,"context_line":"        min\u003d0,"},{"line_number":400,"context_line":"        help\u003d\"\"\""},{"line_number":401,"context_line":"Number of physical CPUs to reserve for the host so that it is always available"},{"line_number":402,"context_line":"to host processes. The host resources usage is reported back to the scheduler"},{"line_number":403,"context_line":"continuously from nova-compute running on the compute node. To prevent the host"},{"line_number":404,"context_line":"CPU from being considered as available, this option is used to reserve"},{"line_number":405,"context_line":"pCPU(s) for the host."}],"source_content_type":"text/x-python","patch_set":1,"id":"5ff73747_1088e86e","line":402,"updated":"2017-04-26 14:48:40.000000000","message":"s/so that it is always available to host processes//\n\nThe reason is because host processes are not guaranteed to run only on the reserved pCPU (since, obviously, we\u0027re not reserving a *specific* pCPU).\n\nPlus, it\u0027s just filler and you don\u0027t need it to clarify anything further.","commit_id":"2492ad97b86a4e5a13afcd2fe0f14f635664b69e"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"13eb0dd359ca98f84ea245b9230545d3f83d79ea","unresolved":false,"context_lines":[{"line_number":394,"context_line":"* Any positive integer representing amount of memory in MB to reserve"},{"line_number":395,"context_line":"  for the host."},{"line_number":396,"context_line":"\"\"\"),"},{"line_number":397,"context_line":"    cfg.IntOpt(\u0027reserved_host_cpus\u0027,"},{"line_number":398,"context_line":"        default\u003d0,"},{"line_number":399,"context_line":"        min\u003d0,"},{"line_number":400,"context_line":"        help\u003d\"\"\""}],"source_content_type":"text/x-python","patch_set":2,"id":"5ff73747_3965bf46","line":397,"updated":"2017-04-28 14:27:16.000000000","message":"sfinucan pointed this out in IRC, but we might want to note the related option vcpu_pin_set and probably say something to the effect that if you\u0027re pinning CPUs with vcpu_pin_set then you shouldn\u0027t need to be setting this option either.","commit_id":"c102600d59ed6cd8751d8012c138cd318866910a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"3e930f7fbdd08857e7b8dfe7a012c83b97a85d58","unresolved":false,"context_lines":[{"line_number":398,"context_line":"        default\u003d0,"},{"line_number":399,"context_line":"        min\u003d0,"},{"line_number":400,"context_line":"        help\u003d\"\"\""},{"line_number":401,"context_line":"Number of physical CPUs to reserve for the host. The host resources usage is"},{"line_number":402,"context_line":"reported back to the scheduler continuously from nova-compute running on the"},{"line_number":403,"context_line":"compute node. To prevent the host CPU from being considered as available,"},{"line_number":404,"context_line":"this option is used to reserve random pCPU(s) for the host."},{"line_number":405,"context_line":""},{"line_number":406,"context_line":"Possible values:"},{"line_number":407,"context_line":""},{"line_number":408,"context_line":"* Any positive integer representing number of physical CPUs to reserve"},{"line_number":409,"context_line":"  for the host."},{"line_number":410,"context_line":"\"\"\"),"},{"line_number":411,"context_line":"]"},{"line_number":412,"context_line":""}],"source_content_type":"text/x-python","patch_set":2,"id":"5ff73747_ae062351","line":409,"range":{"start_line":401,"start_character":0,"end_line":409,"end_character":15},"updated":"2017-04-28 14:08:48.000000000","message":"At some point I think we need to explain why this matters wrt the overhead calculation for claims in the scheduler, but we could handle that later when actually make those changes in the compute code. When that change happens we should also use the DocImpact tag to generate a docs bug so we can document the issues with the overhead calculation and give examples of how to do it (Jay had examples about being conservative vs aggressive).","commit_id":"c102600d59ed6cd8751d8012c138cd318866910a"}],"releasenotes/notes/reserved_host_cpus-e7de4aa9b89bd947.yaml":[{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"3e930f7fbdd08857e7b8dfe7a012c83b97a85d58","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"features:"},{"line_number":3,"context_line":"  - A new configuration option ``reserved_host_cpus`` has been added for"},{"line_number":4,"context_line":"    compute services. It helps operators to provide how many physical CPUs they"},{"line_number":5,"context_line":"    would like to reserve for the hypervisor separately from what the instances"},{"line_number":6,"context_line":"    use."}],"source_content_type":"text/x-yaml","patch_set":2,"id":"5ff73747_aeb4e380","line":4,"range":{"start_line":4,"start_character":41,"end_line":4,"end_character":51},"updated":"2017-04-28 14:08:48.000000000","message":"specify","commit_id":"c102600d59ed6cd8751d8012c138cd318866910a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"3e930f7fbdd08857e7b8dfe7a012c83b97a85d58","unresolved":false,"context_lines":[{"line_number":3,"context_line":"  - A new configuration option ``reserved_host_cpus`` has been added for"},{"line_number":4,"context_line":"    compute services. It helps operators to provide how many physical CPUs they"},{"line_number":5,"context_line":"    would like to reserve for the hypervisor separately from what the instances"},{"line_number":6,"context_line":"    use."}],"source_content_type":"text/x-yaml","patch_set":2,"id":"5ff73747_ee442b76","line":6,"updated":"2017-04-28 14:08:48.000000000","message":"Again, at some point later when the compute claim code happens, we\u0027re going to want to expand on the need for this because of the loss of overhead calculations when making a claim (replaced with allocations). But we could do that in the later change that does all of that code in the compute service if you prefer.","commit_id":"c102600d59ed6cd8751d8012c138cd318866910a"}]}