)]}'
{"nova/tests/functional/regressions/test_bug_1790204.py":[{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"9fb9fcbe182599127a7f73a11b0464a99774f6c3","unresolved":false,"context_lines":[{"line_number":70,"context_line":"        # resource provider VCPU inventory is 2 and DISK_GB is 1028, but the"},{"line_number":71,"context_line":"        # doubled up allocations would be VCPU\u003d3 and DISK_GB\u003d2048 even though"},{"line_number":72,"context_line":"        # flavor2 only needs VCPU\u003d2 and DISK_GB\u003d1028 which should fit."},{"line_number":73,"context_line":"        ex \u003d self.assertRaises(api_client.OpenStackApiException,"},{"line_number":74,"context_line":"                               self.api.post_server_action,"},{"line_number":75,"context_line":"                               server[\u0027id\u0027], resize_req)"},{"line_number":76,"context_line":"        self.assertEqual(400, ex.response.status_code)"}],"source_content_type":"text/x-python","patch_set":1,"id":"3f79a3b5_06196d22","line":73,"updated":"2018-12-20 22:04:25.000000000","message":"Hmm, thinking about this, I\u0027m kind of confused why this doesn\u0027t work since this test isn\u0027t setting cpu_allocation_ratio\u003d1.0, so really if the SmallFakeDriver is reporting 2 vcpus, why wouldn\u0027t placement actually be allowing up to 32 (2*16.0) on the host?","commit_id":"f631d0a900f34c373e7684e289336e4620860a50"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"c7a347340a5eb1cde0e5c8f28a4044ffac537b12","unresolved":false,"context_lines":[{"line_number":70,"context_line":"        # resource provider VCPU inventory is 2 and DISK_GB is 1028, but the"},{"line_number":71,"context_line":"        # doubled up allocations would be VCPU\u003d3 and DISK_GB\u003d2048 even though"},{"line_number":72,"context_line":"        # flavor2 only needs VCPU\u003d2 and DISK_GB\u003d1028 which should fit."},{"line_number":73,"context_line":"        ex \u003d self.assertRaises(api_client.OpenStackApiException,"},{"line_number":74,"context_line":"                               self.api.post_server_action,"},{"line_number":75,"context_line":"                               server[\u0027id\u0027], resize_req)"},{"line_number":76,"context_line":"        self.assertEqual(400, ex.response.status_code)"}],"source_content_type":"text/x-python","patch_set":1,"id":"3f79a3b5_2c01ebb4","line":73,"in_reply_to":"3f79a3b5_06196d22","updated":"2018-12-20 22:59:28.000000000","message":"VCPU isn\u0027t the problem that causes NoValidHost, it\u0027s disk:\n\n    b\u00272018-12-20 17:44:38,168 DEBUG [placement.requestlog] Starting request: 127.0.0.1 \"GET /placement/allocation_candidates?limit\u003d1000\u0026resources\u003dDISK_GB%3A1028%2CMEMORY_MB%3A2048%2CVCPU%3A2\"\u0027\n    b\u00272018-12-20 17:44:38,183 DEBUG [placement.objects.resource_provider] found 1 providers with available 2 VCPU\u0027\n    b\u00272018-12-20 17:44:38,186 DEBUG [placement.objects.resource_provider] found 1 providers with available 2048 MEMORY_MB\u0027\n    b\u00272018-12-20 17:44:38,186 DEBUG [placement.objects.resource_provider] found 1 providers after filtering by previous result\u0027\n    b\u00272018-12-20 17:44:38,190 DEBUG [placement.objects.resource_provider] found 0 providers with available 1028 DISK_GB\u0027\n    b\"2018-12-20 17:44:38,190 DEBUG [placement.objects.resource_provider] RequestGroup(use_same_provider\u003dFalse, resources\u003d{DISK_GB:1028, MEMORY_MB:2048, VCPU:2}, traits\u003d[], aggregates\u003d[]) (suffix \u0027\u0027) returned 0 matches\"\n    b\u00272018-12-20 17:44:38,191 INFO [placement.requestlog] 127.0.0.1 \"GET /placement/allocation_candidates?limit\u003d1000\u0026resources\u003dDISK_GB%3A1028%2CMEMORY_MB%3A2048%2CVCPU%3A2\" status: 200 len: 53 microversion: 1.29\u0027\n    b\u00272018-12-20 17:44:38,192 INFO [nova.scheduler.manager] Got no allocation candidates from the Placement API. This could be due to insufficient resources or a temporary occurrence as compute nodes start up.\u0027\n\n\nBecause initial_disk_allocation_ratio\u003d1.0 and we\u0027re trying to claim 2048 on that provider.","commit_id":"f631d0a900f34c373e7684e289336e4620860a50"},{"author":{"_account_id":7,"name":"Jay Pipes","email":"jaypipes@gmail.com","username":"jaypipes"},"change_message_id":"4ad390a14bb830c46f51c425ce8dffdf6903c652","unresolved":false,"context_lines":[{"line_number":67,"context_line":"        # FIXME(mriedem): This is bug 1790204 where scheduling fails with"},{"line_number":68,"context_line":"        # NoValidHost because the scheduler fails to claim \"doubled up\""},{"line_number":69,"context_line":"        # allocations on the same compute node resource provider. Scheduling"},{"line_number":70,"context_line":"        # fails because DISK_GB inventory allocation ratio is 1.0 and when"},{"line_number":71,"context_line":"        # resizing to the same host we\u0027ll be trying to claim 2048 but the"},{"line_number":72,"context_line":"        # provider only has a total of 1024."},{"line_number":73,"context_line":"        ex \u003d self.assertRaises(api_client.OpenStackApiException,"}],"source_content_type":"text/x-python","patch_set":2,"id":"1f769fc5_f54810cd","line":70,"range":{"start_line":70,"start_character":16,"end_line":70,"end_character":65},"updated":"2019-01-02 15:10:00.000000000","message":"yup... good that you mention the role that the allocation ratio plays in this, since it\u0027s not actually the vCPU change that will cause the failure -- and since your flavor2 only changes the vcpus count, that\u0027s a bit misleading.","commit_id":"0ce581369838dbe6909bdd8847aefac2aff1177e"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"d9a39798e88b66c08e355b4099393ae6853d6b33","unresolved":false,"context_lines":[{"line_number":67,"context_line":"        # FIXME(mriedem): This is bug 1790204 where scheduling fails with"},{"line_number":68,"context_line":"        # NoValidHost because the scheduler fails to claim \"doubled up\""},{"line_number":69,"context_line":"        # allocations on the same compute node resource provider. Scheduling"},{"line_number":70,"context_line":"        # fails because DISK_GB inventory allocation ratio is 1.0 and when"},{"line_number":71,"context_line":"        # resizing to the same host we\u0027ll be trying to claim 2048 but the"},{"line_number":72,"context_line":"        # provider only has a total of 1024."},{"line_number":73,"context_line":"        ex \u003d self.assertRaises(api_client.OpenStackApiException,"}],"source_content_type":"text/x-python","patch_set":2,"id":"ffd0ebdf_1252c168","line":70,"range":{"start_line":70,"start_character":16,"end_line":70,"end_character":65},"in_reply_to":"1f769fc5_f54810cd","updated":"2019-01-03 14:25:41.000000000","message":"Yeah, I actually thought it was the VCPU change at first until I dug into this a bit more and realized it was the disk filtering that kicked it out (after running this with debug enabled to see the logging in placement).","commit_id":"0ce581369838dbe6909bdd8847aefac2aff1177e"},{"author":{"_account_id":5754,"name":"Alex Xu","email":"hejie.xu@intel.com","username":"xuhj"},"change_message_id":"c104ef3bd1bedd8406f6124f5e02e5c66b31ede9","unresolved":false,"context_lines":[{"line_number":68,"context_line":"        # NoValidHost because the scheduler fails to claim \"doubled up\""},{"line_number":69,"context_line":"        # allocations on the same compute node resource provider. Scheduling"},{"line_number":70,"context_line":"        # fails because DISK_GB inventory allocation ratio is 1.0 and when"},{"line_number":71,"context_line":"        # resizing to the same host we\u0027ll be trying to claim 2048 but the"},{"line_number":72,"context_line":"        # provider only has a total of 1024."},{"line_number":73,"context_line":"        ex \u003d self.assertRaises(api_client.OpenStackApiException,"},{"line_number":74,"context_line":"                               self.api.post_server_action,"}],"source_content_type":"text/x-python","patch_set":2,"id":"ffd0ebdf_e60c2644","line":71,"range":{"start_line":71,"start_character":61,"end_line":71,"end_character":66},"updated":"2019-01-03 07:47:52.000000000","message":"2056, not sure why we have 1024 in the fake driver.","commit_id":"0ce581369838dbe6909bdd8847aefac2aff1177e"},{"author":{"_account_id":5754,"name":"Alex Xu","email":"hejie.xu@intel.com","username":"xuhj"},"change_message_id":"c104ef3bd1bedd8406f6124f5e02e5c66b31ede9","unresolved":false,"context_lines":[{"line_number":69,"context_line":"        # allocations on the same compute node resource provider. Scheduling"},{"line_number":70,"context_line":"        # fails because DISK_GB inventory allocation ratio is 1.0 and when"},{"line_number":71,"context_line":"        # resizing to the same host we\u0027ll be trying to claim 2048 but the"},{"line_number":72,"context_line":"        # provider only has a total of 1024."},{"line_number":73,"context_line":"        ex \u003d self.assertRaises(api_client.OpenStackApiException,"},{"line_number":74,"context_line":"                               self.api.post_server_action,"},{"line_number":75,"context_line":"                               server[\u0027id\u0027], resize_req)"}],"source_content_type":"text/x-python","patch_set":2,"id":"ffd0ebdf_46189280","line":72,"range":{"start_line":72,"start_character":39,"end_line":72,"end_character":43},"updated":"2019-01-03 07:47:52.000000000","message":"actually, 1028","commit_id":"0ce581369838dbe6909bdd8847aefac2aff1177e"}]}