File size: 28,991 Bytes
e8c001c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
-- mysql_006 database initialization: create tables + load seed data
-- Executed via: mysql -u root < init_db.sql

-- Fix: MySQL root default auth_socket blocks pymysql TCP connections,
-- switch to mysql_native_password so pymysql (Python) can connect.
ALTER USER 'root'@'localhost' IDENTIFIED WITH mysql_native_password BY 'root123';
FLUSH PRIVILEGES;

CREATE DATABASE IF NOT EXISTS internal_platform_db
  DEFAULT CHARACTER SET utf8mb4
  DEFAULT COLLATE utf8mb4_unicode_ci;

USE internal_platform_db;

-- 1. Input table: dwm_gputj_platform_gpu_base_feature_agg_v2
-- Columns: dt, service_name, instance_uuid, agg_type, agg_time, pod_count,
--   dcgm_fi_dev_gpu_util_pod_avg, k8s_container_bs_rate_cpu_core_used_request_pod_avg,
--   k8s_container_rate_mem_working_set_request_pod_avg, k8s_dcgm_fi_dev_fb_util_pod_avg,
--   k8s_container_vgpu_gpu_util_pod_avg,
--   dcgm_fi_dev_gpu_util_pod_max, k8s_container_bs_rate_cpu_core_used_request_pod_max,
--   k8s_container_rate_mem_working_set_request_pod_max, k8s_dcgm_fi_dev_fb_util_pod_max,
--   k8s_container_vgpu_gpu_util_pod_max,
--   dcgm_fi_dev_gpu_util_model_sum,
--   nv_inference_count_model_avg, nv_inference_count_model_max,
--   nv_inference_request_duration_ms_model_avg, nv_inference_request_duration_ms_model_max,
--   nv_inference_queue_duration_ms_model_avg, nv_inference_queue_duration_ms_model_max,
--   num_queued_reqs_model_avg, num_queued_reqs_model_max,
--   nv_inference_request_success_model_avg, nv_inference_request_success_model_max,
--   nv_inference_request_failure_model_avg, nv_inference_request_failure_model_max,
--   nv_inference_request_duration_ms_perreq_avg, nv_inference_request_duration_ms_perreq_max,
--   nv_inference_queue_duration_ms_perreq_avg, nv_inference_queue_duration_ms_perreq_max,
--   nv_inference_request_duration_ms_perreq_p95, nv_inference_queue_duration_ms_perreq_p95,
--   nv_inference_count_pod_avg, nv_inference_count_pod_max,
--   is_elasticity, scene, replicas, task_priority_expect, gpu_res_expect,
--   scale_up_target_size, scale_down_target_size, scale_up_time_str, scale_down_time_str,
--   max_replicas, gpu_name, queue_name, project_name, wsid_tag,
--   scene_type, business_scene, service_scene, queue_module, compression_strategy,
--   host_gpu_num, host_num, avg_pod_count, namespace, trial_job_name,
--   stream_request_total_model_avg, stream_request_total_model_max,
--   stream_request_timeout_total_model_avg, stream_request_timeout_total_model_max,
--   stream_empty_output_total_model_avg, stream_empty_output_total_model_max,
--   stream_request_fail_total_model_avg, stream_request_fail_total_model_max,
--   stream_cost_time_ms_perreq_avg, stream_cost_time_ms_perreq_max,
--   stream_first_char_time_ms_perreq_avg, stream_first_char_time_ms_perreq_max,
--   pod_req_count_mean_abs_rate_of_change, pod_req_count_mean_abs_rate_accelerated_of_change,
--   pod_req_count_coefficient_of_variation, pod_req_count_stddev_value,
--   nv_inference_count_model_min, nv_inference_count_model_p50, nv_inference_count_model_p90,
--   nv_inference_count_model_max_minute, nv_inference_count_model_min_minute,
--   nv_inference_request_duration_ms_perreq_min, nv_inference_request_duration_ms_perreq_p50,
--   nv_inference_request_duration_ms_perreq_p90,
--   nv_inference_request_duration_ms_perreq_max_minute, nv_inference_request_duration_ms_perreq_min_minute,
--   request_count, request_failed_count, request_failed_rate_avg, request_failed_rate_max,
--   request_timeout_count, request_timeout_rate_avg, request_timeout_rate_max,
--   first_char_time_perreq_ms_avg, first_char_time_perreq_ms_max, first_char_time_perreq_ms_min,
--   output_token_rate_model_avg, output_token_rate_model_max,
--   token_interval_avg_model_avg, token_interval_avg_model_max,
--   infer_request_total_model_avg, infer_request_total_model_max,
--   infer_request_timeout_total_model_avg, infer_request_timeout_total_model_max,
--   infer_empty_output_total_model_avg, infer_empty_output_total_model_max,
--   infer_request_fail_total_model_avg, infer_request_fail_total_model_max,
--   infer_cost_time_ms_perreq_avg, infer_cost_time_ms_perreq_max,
--   infer_first_char_time_ms_perreq_avg, infer_first_char_time_ms_perreq_max

CREATE TABLE IF NOT EXISTS dwm_gputj_platform_gpu_base_feature_agg_v2_mysql_006 (
    dt VARCHAR(256),
    service_name VARCHAR(256),
    instance_uuid VARCHAR(256),
    agg_type BIGINT,
    agg_time VARCHAR(256),
    pod_count BIGINT,
    dcgm_fi_dev_gpu_util_pod_avg DOUBLE,
    k8s_container_bs_rate_cpu_core_used_request_pod_avg DOUBLE,
    k8s_container_rate_mem_working_set_request_pod_avg DOUBLE,
    k8s_dcgm_fi_dev_fb_util_pod_avg DOUBLE,
    k8s_container_vgpu_gpu_util_pod_avg DOUBLE,
    dcgm_fi_dev_gpu_util_pod_max DOUBLE,
    k8s_container_bs_rate_cpu_core_used_request_pod_max DOUBLE,
    k8s_container_rate_mem_working_set_request_pod_max DOUBLE,
    k8s_dcgm_fi_dev_fb_util_pod_max DOUBLE,
    k8s_container_vgpu_gpu_util_pod_max DOUBLE,
    dcgm_fi_dev_gpu_util_model_sum DOUBLE,
    nv_inference_count_model_avg DOUBLE,
    nv_inference_count_model_max DOUBLE,
    nv_inference_request_duration_ms_model_avg DOUBLE,
    nv_inference_request_duration_ms_model_max DOUBLE,
    nv_inference_queue_duration_ms_model_avg DOUBLE,
    nv_inference_queue_duration_ms_model_max DOUBLE,
    num_queued_reqs_model_avg DOUBLE,
    num_queued_reqs_model_max DOUBLE,
    nv_inference_request_success_model_avg DOUBLE,
    nv_inference_request_success_model_max DOUBLE,
    nv_inference_request_failure_model_avg DOUBLE,
    nv_inference_request_failure_model_max DOUBLE,
    nv_inference_request_duration_ms_perreq_avg DOUBLE,
    nv_inference_request_duration_ms_perreq_max DOUBLE,
    nv_inference_queue_duration_ms_perreq_avg DOUBLE,
    nv_inference_queue_duration_ms_perreq_max DOUBLE,
    nv_inference_request_duration_ms_perreq_p95 DOUBLE,
    nv_inference_queue_duration_ms_perreq_p95 DOUBLE,
    nv_inference_count_pod_avg DOUBLE,
    nv_inference_count_pod_max DOUBLE,
    is_elasticity BIGINT,
    scene VARCHAR(256),
    replicas BIGINT,
    task_priority_expect BIGINT,
    gpu_res_expect DOUBLE,
    scale_up_target_size BIGINT,
    scale_down_target_size BIGINT,
    scale_up_time_str VARCHAR(256),
    scale_down_time_str VARCHAR(256),
    max_replicas BIGINT,
    gpu_name VARCHAR(256),
    queue_name VARCHAR(256),
    project_name VARCHAR(256),
    wsid_tag VARCHAR(256),
    scene_type VARCHAR(256),
    business_scene VARCHAR(256),
    service_scene VARCHAR(256),
    queue_module VARCHAR(256),
    compression_strategy VARCHAR(256),
    host_gpu_num BIGINT,
    host_num BIGINT,
    avg_pod_count DOUBLE,
    namespace VARCHAR(256),
    trial_job_name VARCHAR(256),
    stream_request_total_model_avg DOUBLE,
    stream_request_total_model_max DOUBLE,
    stream_request_timeout_total_model_avg DOUBLE,
    stream_request_timeout_total_model_max DOUBLE,
    stream_empty_output_total_model_avg DOUBLE,
    stream_empty_output_total_model_max DOUBLE,
    stream_request_fail_total_model_avg DOUBLE,
    stream_request_fail_total_model_max DOUBLE,
    stream_cost_time_ms_perreq_avg DOUBLE,
    stream_cost_time_ms_perreq_max DOUBLE,
    stream_first_char_time_ms_perreq_avg DOUBLE,
    stream_first_char_time_ms_perreq_max DOUBLE,
    pod_req_count_mean_abs_rate_of_change DOUBLE,
    pod_req_count_mean_abs_rate_accelerated_of_change DOUBLE,
    pod_req_count_coefficient_of_variation DOUBLE,
    pod_req_count_stddev_value DOUBLE,
    nv_inference_count_model_min DOUBLE,
    nv_inference_count_model_p50 DOUBLE,
    nv_inference_count_model_p90 DOUBLE,
    nv_inference_count_model_max_minute BIGINT,
    nv_inference_count_model_min_minute BIGINT,
    nv_inference_request_duration_ms_perreq_min DOUBLE,
    nv_inference_request_duration_ms_perreq_p50 DOUBLE,
    nv_inference_request_duration_ms_perreq_p90 DOUBLE,
    nv_inference_request_duration_ms_perreq_max_minute BIGINT,
    nv_inference_request_duration_ms_perreq_min_minute BIGINT,
    request_count DOUBLE,
    request_failed_count DOUBLE,
    request_failed_rate_avg DOUBLE,
    request_failed_rate_max DOUBLE,
    request_timeout_count DOUBLE,
    request_timeout_rate_avg DOUBLE,
    request_timeout_rate_max DOUBLE,
    first_char_time_perreq_ms_avg DOUBLE,
    first_char_time_perreq_ms_max DOUBLE,
    first_char_time_perreq_ms_min DOUBLE,
    output_token_rate_model_avg DOUBLE,
    output_token_rate_model_max DOUBLE,
    token_interval_avg_model_avg DOUBLE,
    token_interval_avg_model_max DOUBLE,
    infer_request_total_model_avg DOUBLE,
    infer_request_total_model_max DOUBLE,
    infer_request_timeout_total_model_avg DOUBLE,
    infer_request_timeout_total_model_max DOUBLE,
    infer_empty_output_total_model_avg DOUBLE,
    infer_empty_output_total_model_max DOUBLE,
    infer_request_fail_total_model_avg DOUBLE,
    infer_request_fail_total_model_max DOUBLE,
    infer_cost_time_ms_perreq_avg DOUBLE,
    infer_cost_time_ms_perreq_max DOUBLE,
    infer_first_char_time_ms_perreq_avg DOUBLE,
    infer_first_char_time_ms_perreq_max DOUBLE
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;

-- Helper: 88-column row template for feature table
-- (dt, service_name, instance_uuid, agg_type, agg_time, pod_count,
--  5 pod_avg, 5 pod_max, 1 model_sum,
--  2 inference_count, 2 req_duration_model, 2 queue_duration_model,
--  2 queued_reqs, 2 req_success, 2 req_failure,
--  2 duration_perreq, 2 queue_perreq, 2 perreq_p95,
--  2 count_pod,
--  is_elasticity, scene, replicas, task_priority_expect, gpu_res_expect,
--  scale_up_target_size, scale_down_target_size, scale_up_time_str, scale_down_time_str,
--  max_replicas, gpu_name, queue_name, project_name, wsid_tag,
--  scene_type, business_scene, service_scene, queue_module, compression_strategy,
--  host_gpu_num, host_num, avg_pod_count, namespace, trial_job_name,
--  12 stream_*, 4 pod_req_count_*, 3 nv_count_min/p50/p90, 2 max/min_minute,
--  3 perreq_min/p50/p90, 2 perreq_max/min_minute,
--  7 request_*, 3 first_char_*, 2 output_token_*, 2 token_interval_*,
--  12 infer_*)

-- svc_alpha, agg_type=1: 6 rows (5 weekday + 1 holiday)
-- Holiday dates in history: 2026-04-27
-- Weekday dates: 2026-04-22, 2026-04-23, 2026-04-24, 2026-04-25, 2026-04-28
INSERT INTO dwm_gputj_platform_gpu_base_feature_agg_v2_mysql_006 VALUES
-- svc_alpha agg_type=1 weekday rows (nv_inference_count_model_avg ascending for P90 ranking)
('20260429', 'svc_alpha', 'uuid-001', 1, '2026-04-22 10:00:00', 2,
 40.0, 25.0, 35.0, 15.0, 45.0, 50.0, 30.0, 40.0, 20.0, 50.0, 80.0,
 80.0, 100.0, 60.0, 80.0, 8.0, 12.0, 3.0, 5.0, 100.0, 120.0, 1.0, 2.0,
 55.0, 75.0, 6.0, 10.0, 60.0, 7.0, 50.0, 70.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 70,80,90,8,3, 50,55,60,10,2, 400,3,0.005,0.02,1,0.002,0.005, 40,70,25, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260430', 'svc_alpha', 'uuid-001', 1, '2026-04-23 10:00:00', 2,
 45.0, 27.0, 37.0, 17.0, 50.0, 55.0, 32.0, 42.0, 22.0, 55.0, 85.0,
 90.0, 110.0, 65.0, 85.0, 9.0, 13.0, 4.0, 6.0, 105.0, 125.0, 1.5, 2.5,
 60.0, 80.0, 7.0, 11.0, 65.0, 8.0, 55.0, 75.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 85,95,105,9,4, 55,60,65,11,3, 420,4,0.006,0.03,1.5,0.003,0.006, 45,75,27, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260501', 'svc_alpha', 'uuid-001', 1, '2026-04-24 10:00:00', 2,
 50.0, 30.0, 40.0, 20.0, 55.0, 60.0, 35.0, 45.0, 25.0, 60.0, 100.0,
 110.0, 130.0, 75.0, 95.0, 10.0, 14.0, 5.0, 7.0, 110.0, 130.0, 2.0, 3.0,
 68.0, 88.0, 8.0, 12.0, 73.0, 9.0, 60.0, 80.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 100,110,120,10,5, 60,65,70,12,3, 450,4.5,0.008,0.04,2,0.004,0.008, 50,80,30, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260502', 'svc_alpha', 'uuid-001', 1, '2026-04-25 10:00:00', 2,
 50.0, 30.0, 40.0, 20.0, 55.0, 60.0, 35.0, 45.0, 25.0, 60.0, 100.0,
 120.5, 150.0, 80.0, 100.0, 10.0, 15.0, 5.0, 8.0, 110.0, 130.0, 2.0, 3.0,
 70.0, 90.0, 8.0, 12.0, 75.0, 9.0, 60.0, 80.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 100,110,130,10,5, 60,65,72,12,3, 500,5,0.01,0.05,2,0.004,0.01, 50,80,30, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260506', 'svc_alpha', 'uuid-001', 1, '2026-04-28 10:00:00', 2,
 55.0, 33.0, 43.0, 23.0, 60.0, 65.0, 38.0, 48.0, 28.0, 65.0, 110.0,
 140.0, 170.0, 90.0, 110.0, 14.0, 18.0, 7.0, 10.0, 120.0, 140.0, 3.5, 4.5,
 75.0, 95.0, 10.0, 14.0, 80.0, 11.0, 65.0, 85.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 130,140,160,12,7, 65,70,76,14,5, 550,6,0.012,0.06,3,0.006,0.012, 55,85,35, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
-- svc_alpha agg_type=1 holiday row (2026-04-27 is a holiday)
('20260504', 'svc_alpha', 'uuid-001', 1, '2026-04-27 10:00:00', 2,
 52.0, 32.0, 42.0, 22.0, 57.0, 62.0, 37.0, 47.0, 27.0, 62.0, 105.0,
 130.0, 160.0, 85.0, 105.0, 12.0, 17.0, 6.0, 9.0, 115.0, 135.0, 3.0, 4.0,
 72.0, 92.0, 9.0, 13.0, 77.0, 10.0, 62.0, 82.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 120,130,150,11,6, 62,67,74,13,4, 520,6,0.012,0.06,3,0.005,0.012, 52,82,32, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),

-- svc_alpha, agg_type=2: 4 rows (3 weekday + 1 holiday)
-- Weekday: 2026-04-22, 2026-04-24, 2026-04-28; Holiday: 2026-04-27
('20260429', 'svc_alpha', 'uuid-001', 2, '2026-04-22 11:00:00', 2,
 38.0, 23.0, 33.0, 13.0, 43.0, 48.0, 28.0, 38.0, 18.0, 48.0, 75.0,
 70.0, 90.0, 50.0, 70.0, 6.0, 10.0, 2.0, 4.0, 90.0, 110.0, 0.5, 1.5,
 50.0, 70.0, 5.0, 9.0, 55.0, 6.0, 45.0, 65.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 60,70,80,7,2, 45,50,55,9,1, 350,2,0.004,0.015,0.5,0.001,0.004, 35,65,20, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260501', 'svc_alpha', 'uuid-001', 2, '2026-04-24 11:00:00', 2,
 48.0, 28.0, 38.0, 18.0, 53.0, 58.0, 33.0, 43.0, 23.0, 58.0, 95.0,
 100.0, 120.0, 70.0, 90.0, 8.0, 12.0, 4.0, 6.0, 105.0, 125.0, 1.5, 2.5,
 65.0, 85.0, 7.0, 11.0, 70.0, 8.0, 55.0, 75.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 95,105,115,10,4, 58,63,68,11,2, 480,4,0.008,0.035,1.5,0.003,0.008, 48,78,28, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260506', 'svc_alpha', 'uuid-001', 2, '2026-04-28 11:00:00', 2,
 52.0, 32.0, 42.0, 22.0, 57.0, 62.0, 37.0, 47.0, 27.0, 62.0, 105.0,
 125.0, 155.0, 82.0, 102.0, 12.0, 16.0, 6.0, 9.0, 115.0, 135.0, 3.0, 4.0,
 72.0, 92.0, 9.0, 13.0, 77.0, 10.0, 60.0, 80.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 120,130,150,12,6, 65,70,76,13,4, 530,5.5,0.011,0.055,2.5,0.005,0.011, 52,82,32, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
-- svc_alpha agg_type=2 holiday row
('20260504', 'svc_alpha', 'uuid-001', 2, '2026-04-27 11:00:00', 2,
 48.0, 28.0, 38.0, 18.0, 53.0, 58.0, 33.0, 43.0, 23.0, 58.0, 98.0,
 100.0, 140.0, 75.0, 95.0, 8.0, 13.0, 4.0, 7.0, 105.0, 125.0, 1.0, 2.0,
 68.0, 88.0, 7.0, 11.0, 73.0, 8.0, 58.0, 78.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q1', 'proj1', 'wsid1', 'llm', 'bs1', 'ss1', 'qm1', 'none',
 8, 1, 3.0, 'ns_test', 'wl_alpha',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 95,105,125,9,4, 58,63,70,11,2, 480,4,0.008,0.04,1,0.003,0.008, 48,78,28, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),

-- svc_beta, agg_type=1: 3 rows (ALL weekday, NO holiday => triggers degradation for holiday predictions)
('20260430', 'svc_beta', 'uuid-002', 1, '2026-04-23 14:00:00', 1,
 35.0, 20.0, 30.0, 10.0, 40.0, 45.0, 25.0, 35.0, 15.0, 45.0, 60.0,
 60.0, 80.0, 45.0, 65.0, 5.0, 8.0, 2.0, 3.0, 85.0, 105.0, 0.5, 1.0,
 45.0, 65.0, 4.0, 7.0, 50.0, 5.0, 40.0, 60.0,
 0, 'inference', 2, 1, 1.5, 3, 1, '09:00', '21:00', 8,
 'V100', 'q2', 'proj2', 'wsid2', 'llm', 'bs2', 'ss2', 'qm2', 'none',
 4, 1, 2.0, 'ns_prod', 'wl_beta',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 50,60,70,6,1, 40,45,50,8,1, 300,1.5,0.003,0.01,0.3,0.001,0.003, 30,60,15, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260501', 'svc_beta', 'uuid-002', 1, '2026-04-24 14:00:00', 1,
 40.0, 25.0, 35.0, 15.0, 45.0, 50.0, 30.0, 40.0, 20.0, 50.0, 70.0,
 80.0, 100.0, 55.0, 75.0, 7.0, 10.0, 3.0, 4.0, 95.0, 115.0, 0.8, 1.5,
 50.0, 70.0, 5.0, 8.0, 55.0, 6.0, 45.0, 65.0,
 0, 'inference', 2, 1, 1.5, 3, 1, '09:00', '21:00', 8,
 'V100', 'q2', 'proj2', 'wsid2', 'llm', 'bs2', 'ss2', 'qm2', 'none',
 4, 1, 2.0, 'ns_prod', 'wl_beta',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 70,80,90,8,2, 45,50,55,9,2, 350,2,0.004,0.015,0.5,0.001,0.004, 35,65,20, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260506', 'svc_beta', 'uuid-002', 1, '2026-04-28 14:00:00', 1,
 45.0, 28.0, 38.0, 18.0, 50.0, 55.0, 33.0, 43.0, 23.0, 55.0, 80.0,
 100.0, 120.0, 65.0, 85.0, 9.0, 12.0, 4.0, 5.0, 105.0, 125.0, 1.2, 2.0,
 55.0, 75.0, 6.0, 9.0, 60.0, 7.0, 50.0, 70.0,
 0, 'inference', 2, 1, 1.5, 3, 1, '09:00', '21:00', 8,
 'V100', 'q2', 'proj2', 'wsid2', 'llm', 'bs2', 'ss2', 'qm2', 'none',
 4, 1, 2.0, 'ns_prod', 'wl_beta',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 90,100,110,9,3, 50,55,60,10,3, 400,3,0.005,0.02,0.8,0.002,0.005, 40,70,25, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),

-- svc_beta, agg_type=2: 3 rows (ALL weekday, NO holiday => triggers degradation for holiday predictions)
('20260429', 'svc_beta', 'uuid-002', 2, '2026-04-22 14:00:00', 1,
 32.0, 18.0, 28.0, 10.0, 37.0, 42.0, 23.0, 33.0, 15.0, 42.0, 55.0,
 55.0, 75.0, 40.0, 60.0, 4.0, 7.0, 1.5, 2.5, 80.0, 100.0, 0.3, 0.8,
 42.0, 62.0, 3.5, 6.5, 47.0, 4.5, 37.0, 57.0,
 0, 'inference', 2, 1, 1.5, 3, 1, '09:00', '21:00', 8,
 'V100', 'q2', 'proj2', 'wsid2', 'llm', 'bs2', 'ss2', 'qm2', 'none',
 4, 1, 2.0, 'ns_prod', 'wl_beta',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 45,55,65,5,1, 35,40,45,7,1, 280,1,0.002,0.008,0.2,0.001,0.002, 25,55,10, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260502', 'svc_beta', 'uuid-002', 2, '2026-04-25 14:00:00', 1,
 38.0, 23.0, 33.0, 13.0, 43.0, 48.0, 28.0, 38.0, 18.0, 48.0, 65.0,
 75.0, 95.0, 50.0, 70.0, 6.0, 9.0, 2.5, 3.5, 90.0, 110.0, 0.6, 1.2,
 48.0, 68.0, 4.5, 7.5, 52.0, 5.5, 42.0, 62.0,
 0, 'inference', 2, 1, 1.5, 3, 1, '09:00', '21:00', 8,
 'V100', 'q2', 'proj2', 'wsid2', 'llm', 'bs2', 'ss2', 'qm2', 'none',
 4, 1, 2.0, 'ns_prod', 'wl_beta',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 65,75,85,7,2, 42,47,52,9,2, 320,1.8,0.003,0.012,0.4,0.001,0.003, 32,62,17, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
('20260506', 'svc_beta', 'uuid-002', 2, '2026-04-28 14:00:00', 1,
 42.0, 26.0, 36.0, 16.0, 48.0, 52.0, 31.0, 41.0, 21.0, 52.0, 75.0,
 90.0, 110.0, 58.0, 78.0, 8.0, 11.0, 3.5, 4.5, 100.0, 120.0, 1.0, 1.8,
 52.0, 72.0, 5.5, 8.5, 57.0, 6.5, 47.0, 67.0,
 0, 'inference', 2, 1, 1.5, 3, 1, '09:00', '21:00', 8,
 'V100', 'q2', 'proj2', 'wsid2', 'llm', 'bs2', 'ss2', 'qm2', 'none',
 4, 1, 2.0, 'ns_prod', 'wl_beta',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 80,90,100,9,3, 48,53,58,10,3, 370,2.5,0.004,0.018,0.7,0.002,0.004, 38,68,22, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0);

-- ============================================================================
-- Coverage augmentation (诱饵 + 正例) for rules left UNCOVERED by 2-service seed.
-- svc_gamma: agg_type=1, time-part 16:00, history is HOLIDAY-ONLY (2026-04-27).
--   正例 (R8d): weekday predictions have feature_weekday_total=0 (<1) => degrade to
--   feature_holiday_index branch. This is the only service that exercises that branch.
--   Correct CPD breakpoint = '2026-04-26 16:00:00' (lowest metric_value on latest dt).
-- All decoys below are scoped to svc_gamma / the empty-name service, so svc_alpha and
-- svc_beta output rows are completely untouched.
INSERT INTO dwm_gputj_platform_gpu_base_feature_agg_v2_mysql_006 VALUES
-- 正例 G1: svc_gamma holiday history row (2026-04-27, post-breakpoint, the ONLY valid row)
('20260504', 'svc_gamma', 'uuid-003', 1, '2026-04-27 16:00:00', 2,
 66.0, 44.0, 54.0, 34.0, 70.0, 70.0, 50.0, 60.0, 40.0, 75.0, 110.0,
 200.0, 210.0, 95.0, 115.0, 15.0, 20.0, 8.0, 11.0, 130.0, 150.0, 5.0, 6.0,
 85.0, 105.0, 12.0, 16.0, 90.0, 13.0, 70.0, 90.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q3', 'proj3', 'wsid3', 'llm', 'bs3', 'ss3', 'qm3', 'none',
 8, 1, 3.0, 'ns_gamma', 'wl_gamma',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 180,190,200,12,7, 80,85,90,13,4, 600,7,0.012,0.06,3,0.005,0.012, 66,90,40, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
-- 诱饵 R7 (cpd inner dedup): pre-breakpoint weekday row (2026-04-20). Filtered by post-breakpoint
-- join when correct breakpoint (04-26) is picked; LEAKS if dedup flips ASC->DESC (breakpoint 04-15).
('20260422', 'svc_gamma', 'uuid-003', 1, '2026-04-20 16:00:00', 2,
 60.0, 40.0, 50.0, 30.0, 65.0, 65.0, 45.0, 55.0, 35.0, 70.0, 100.0,
 999.0, 1000.0, 90.0, 110.0, 14.0, 19.0, 7.0, 10.0, 125.0, 145.0, 4.0, 5.0,
 80.0, 100.0, 11.0, 15.0, 85.0, 12.0, 65.0, 85.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q3', 'proj3', 'wsid3', 'llm', 'bs3', 'ss3', 'qm3', 'none',
 8, 1, 3.0, 'ns_gamma', 'wl_gamma',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 180,190,200,12,7, 80,85,90,13,4, 600,7,0.012,0.06,3,0.005,0.012, 60,85,35, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
-- 诱饵 R1 (nv_inference_count_model_avg >= 0): negative-avg row (post-breakpoint, weekday).
-- Filtered by the >=0 rule; LEAKS into svc_gamma group if that filter is dropped.
('20260506', 'svc_gamma', 'uuid-003', 1, '2026-04-28 16:00:00', 2,
 60.0, 40.0, 50.0, 30.0, 65.0, 65.0, 45.0, 55.0, 35.0, 70.0, 100.0,
 -10.0, 0.0, 90.0, 110.0, 14.0, 19.0, 7.0, 10.0, 125.0, 145.0, 4.0, 5.0,
 80.0, 100.0, 11.0, 15.0, 85.0, 12.0, 65.0, 85.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q3', 'proj3', 'wsid3', 'llm', 'bs3', 'ss3', 'qm3', 'none',
 8, 1, 3.0, 'ns_gamma', 'wl_gamma',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 180,190,200,12,7, 80,85,90,13,4, 600,7,0.012,0.06,3,0.005,0.012, 60,85,35, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
-- 诱饵 R2 (agg_type IN (1,2)): agg_type=3 row in template window (creates a 16:00/type3 future
-- slot). Filtered out of the feature CTE; LEAKS as type-3 output rows if the agg_type filter widens.
('20260506', 'svc_gamma', 'uuid-003', 3, '2026-04-29 16:00:00', 2,
 60.0, 40.0, 50.0, 30.0, 65.0, 65.0, 45.0, 55.0, 35.0, 70.0, 100.0,
 50.0, 60.0, 90.0, 110.0, 14.0, 19.0, 7.0, 10.0, 125.0, 145.0, 4.0, 5.0,
 80.0, 100.0, 11.0, 15.0, 85.0, 12.0, 65.0, 85.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q3', 'proj3', 'wsid3', 'llm', 'bs3', 'ss3', 'qm3', 'none',
 8, 1, 3.0, 'ns_gamma', 'wl_gamma',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 180,190,200,12,7, 80,85,90,13,4, 600,7,0.012,0.06,3,0.005,0.012, 60,85,35, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0),
-- 诱饵 R4 (service name <> ''): empty-name service with a full valid history row.
-- Excluded from service_list by the name<>'' rule; LEAKS ~14 output rows if that filter is dropped.
('20260504', '', 'uuid-004', 1, '2026-04-27 16:00:00', 2,
 60.0, 40.0, 50.0, 30.0, 65.0, 65.0, 45.0, 55.0, 35.0, 70.0, 100.0,
 300.0, 310.0, 90.0, 110.0, 14.0, 19.0, 7.0, 10.0, 125.0, 145.0, 4.0, 5.0,
 80.0, 100.0, 11.0, 15.0, 85.0, 12.0, 65.0, 85.0,
 0, 'inference', 3, 1, 2.0, 5, 2, '08:00', '22:00', 10,
 'A100', 'q4', 'proj4', 'wsid4', 'llm', 'bs4', 'ss4', 'qm4', 'none',
 8, 1, 3.0, 'ns_empty', 'wl_empty',
 0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0, 180,190,200,12,7, 80,85,90,13,4, 600,7,0.012,0.06,3,0.005,0.012, 60,85,35, 0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0);

-- 2. Input table: dwd_aide_inferencev2_done_service_info_h
CREATE TABLE IF NOT EXISTS dwd_aide_inferencev2_done_service_info_h_mysql_006 (
    id VARCHAR(256),
    name VARCHAR(256),
    dt VARCHAR(256)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;

INSERT INTO dwd_aide_inferencev2_done_service_info_h_mysql_006 VALUES
('1', 'svc_alpha', '2026050723'),
('2', 'svc_beta', '2026050723'),
-- 正例: svc_gamma is an online service => must appear in service_list
('3', 'svc_gamma', '2026050723'),
-- 诱饵 R4 (name <> ''): empty-name service must be excluded from service_list
('4', '', '2026050723');

-- 3. Input table: dwd_gputj_platform_metric_cpd_offline_v2
-- CPD breakpoint dates must be VARCHAR matching the format 'YYYY-MM-DD HH:MM:SS'
-- to avoid implicit type conversion with feature.agg_time
CREATE TABLE IF NOT EXISTS dwd_gputj_platform_metric_cpd_offline_v2_mysql_006 (
    dt VARCHAR(256),
    service_name VARCHAR(256),
    agg_time VARCHAR(256),
    metric_name VARCHAR(256),
    metric_value DOUBLE,
    msg VARCHAR(256)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;

INSERT INTO dwd_gputj_platform_metric_cpd_offline_v2_mysql_006 VALUES
-- svc_alpha: two metric_names on same dt, lowest metric_value per metric wins,
-- then pick latest dt's agg_time -> picks 2026-04-24 10:00:00
('20260423', 'svc_alpha', '2026-04-23 10:00:00', 'cpd_metric', 0.9, 'success'),
('20260423', 'svc_alpha', '2026-04-23 10:00:00', 'cpd_metric2', 0.7, 'success'),
('20260424', 'svc_alpha', '2026-04-24 10:00:00', 'cpd_metric', 0.5, 'success'),
('20260424', 'svc_alpha', '2026-04-24 10:00:00', 'cpd_metric2', 0.6, 'success'),
-- svc_beta: single breakpoint at 2026-04-22
('20260422', 'svc_beta', '2026-04-22 14:00:00', 'cpd_metric', 0.4, 'success'),
-- svc_gamma: correct breakpoint = '2026-04-26 16:00:00'.
--   诱饵 R7 (inner dedup ORDER BY metric_value ASC): same (svc_gamma, dt=20260426, cpd_metric)
--   appears twice. Lowest value (0.3 -> agg '2026-04-26') wins under ASC. If dedup flips to DESC,
--   the 0.9 row (agg '2026-04-29 16:00:00') wins => breakpoint moves past svc_gamma's only history
--   record (2026-04-27) => svc_gamma rows vanish => output diverges.
('20260426', 'svc_gamma', '2026-04-26 16:00:00', 'cpd_metric', 0.3, 'success'),
('20260426', 'svc_gamma', '2026-04-29 16:00:00', 'cpd_metric', 0.9, 'success'),
--   诱饵 R3 (msg = 'success'): a non-success row at a LATER dt with the lowest value. If the
--   msg filter is dropped, this becomes the latest breakpoint (2026-04-30) and drops svc_gamma.
('20260430', 'svc_gamma', '2026-04-30 16:00:00', 'cpd_metric', 0.1, 'failed'),
-- 诱饵 R4 (service name <> ''): breakpoint for the empty-name service so that its feature row
-- survives the post-breakpoint join. The empty service is still excluded by name<>''; if that
-- filter is dropped, it leaks ~14 output rows.
('20260426', '', '2026-04-26 16:00:00', 'cpd_metric', 0.3, 'success');

-- 4. Input table: dim_holiday_list
-- Expanded holidays to cover consecutive holidays and more boundary cases
-- Future window: 2026-05-08 (holiday), 2026-05-09 (holiday) -> consecutive holidays
-- History window: 2026-04-27 (holiday) -> only historical holiday
CREATE TABLE IF NOT EXISTS dim_holiday_list_mysql_006 (
    holiday_date VARCHAR(256)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;

INSERT INTO dim_holiday_list_mysql_006 VALUES
('2026-04-27'),
('2026-05-08'),
('2026-05-09');

-- 5. Output table: dwm_gputj_platform_model_prediction_long_cpd
CREATE TABLE IF NOT EXISTS dwm_gputj_platform_model_prediction_long_cpd_mysql_006 (
    instance_uuid VARCHAR(256),
    service_name VARCHAR(256),
    workload_name VARCHAR(256),
    namespace VARCHAR(256),
    agg_time VARCHAR(256),
    agg_type BIGINT,
    is_holiday BIGINT,
    day_of_week BIGINT,
    prediction_type VARCHAR(256),
    nv_inference_count_model_avg_p90 DOUBLE,
    statistic_time_count BIGINT,
    nv_inference_request_duration_ms_model_avg DOUBLE,
    nv_inference_queue_duration_ms_model_avg DOUBLE,
    num_queued_reqs_model_avg DOUBLE,
    nv_inference_request_success_model_avg DOUBLE,
    nv_inference_request_failure_model_avg DOUBLE,
    nv_inference_request_duration_ms_perreq_avg DOUBLE,
    nv_inference_queue_duration_ms_perreq_avg DOUBLE,
    nv_inference_request_duration_ms_perreq_p95 DOUBLE,
    nv_inference_queue_duration_ms_perreq_p95 DOUBLE,
    nv_inference_request_success_model_max DOUBLE,
    nv_inference_request_failure_model_max DOUBLE,
    DCGM_FI_DEV_GPU_UTIL_pod_avg DOUBLE,
    k8s_container_bs_rate_cpu_core_used_request_pod_avg DOUBLE,
    k8s_container_rate_mem_working_set_request_pod_avg DOUBLE,
    k8s_dcgm_fi_dev_fb_util_pod_avg DOUBLE,
    k8s_container_vgpu_gpu_util_pod_avg DOUBLE,
    cpd_agg_time VARCHAR(256),
    feature_total BIGINT,
    feature_holiday_total BIGINT,
    dt VARCHAR(256)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;