Hugging Face's logo

Datasets:
dicemy
/
DataClawEval

Tasks:
Text Generation
Modalities:
Text
Formats:
json
Languages:
English
Chinese
Size:
< 1K
ArXiv:
Tags:
data-engineering
benchmark
agent
sql
text2sql
pyspark
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
DataClawEval
File size: 11,583 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
-- mysql_014 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;

-- ==================== Input table 1: notebook_span_info ====================
CREATE TABLE IF NOT EXISTS notebook_span_info_mysql_014 (
    `databus_imp_date` VARCHAR(256) COMMENT 'primary partition',
    `trace_id` VARCHAR(256) COMMENT 'trace_id',
    `span_kind` VARCHAR(256) COMMENT 'span_kind',
    `error_code` VARCHAR(256) COMMENT 'error_code',
    `service_name` VARCHAR(256) COMMENT 'service_name',
    `message` VARCHAR(256) COMMENT 'message',
    `parent_span_id` VARCHAR(256) COMMENT 'parent_span_id',
    `span_name` VARCHAR(256) COMMENT 'span_name',
    `duration` BIGINT COMMENT 'duration',
    `span_id` VARCHAR(256) COMMENT 'span_id',
    `space_id` VARCHAR(256) COMMENT 'space_id',
    `span_type` VARCHAR(256) COMMENT 'span_type',
    `start_time` VARCHAR(256) COMMENT 'start_time',
    `end_time` VARCHAR(256) COMMENT 'end_time',
    `service_instance` VARCHAR(256) COMMENT 'service_instance',
    `status_code` INT COMMENT 'status_code',
    `datawd_project_id` VARCHAR(256) COMMENT 'datawd_project_id',
    `datawd_task_id` VARCHAR(256) COMMENT 'datawd_task_id',
    `datawd_task_instance_id` VARCHAR(256) COMMENT 'datawd_task_instance_id',
    `compute_type` VARCHAR(256) COMMENT 'compute_type'
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;

-- Mock data for notebook_span_info
-- trace_id=T001: a complete ray runner.execute flow spanning within same day
-- trace_id=T002: another complete flow, spans across 2 days
INSERT INTO notebook_span_info_mysql_014 VALUES
('2026050700', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'runner.execute', 5000, 'S001', 'SP1', 'SPAN', '1746612000000', '1746612300000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
('2026050700', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'execute.code', 3000, 'S002', 'SP1', 'SPAN', '1746612060000', '1746612240000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
('2026050700', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'runner.killed', 90000, 'S002k', 'SP1', 'SPAN', '1746612180000', '1746612270000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
('2026050700', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'set.permanent.compute', 1000, 'S003', 'SP1', 'SPAN', '1746612010000', '1746612050000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
('2026050400', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'runner.execute', 5000, 'S001b', 'SP1', 'SPAN', '1746612000000', '1746612300000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
('2026050400', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'runner.killed', 90000, 'S002kb', 'SP1', 'SPAN', '1746612180000', '1746612270000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
('2026050700', 'T002', 'SERVER', '0', 'notebook-runner', '', '', 'runner.execute', 90000, 'S010', 'SP2', 'SPAN', '1746525600000', '1746612060000', 'inst2', 1, 'P200', 'TASK200', 'INST200', 'ray'),
('2026050400', 'T002', 'SERVER', '0', 'notebook-runner', '', '', 'runner.execute', 90000, 'S010b', 'SP2', 'SPAN', '1746525600000', '1746612060000', 'inst2', 1, 'P200', 'TASK200', 'INST200', 'ray'),
('2026050700', 'T002', 'SERVER', '0', 'notebook-runner', '', '', 'execute.code', 80000, 'S011', 'SP2', 'SPAN', '1746525660000', '1746612000000', 'inst2', 1, 'P200', 'TASK200', 'INST200', 'ray'),
('2026050400', 'T002', 'SERVER', '0', 'notebook-runner', '', '', 'execute.code', 80000, 'S011b', 'SP2', 'SPAN', '1746525660000', '1746612000000', 'inst2', 1, 'P200', 'TASK200', 'INST200', 'ray'),
('2026050700', 'T002', 'SERVER', '0', 'notebook-runner', '', '', 'create.non.permanent.compute', 2000, 'S012', 'SP2', 'SPAN', '1746525610000', '1746525650000', 'inst2', 1, 'P200', 'TASK200', 'INST200', 'ray'),
('2026050400', 'T002', 'SERVER', '0', 'notebook-runner', '', '', 'create.non.permanent.compute', 2000, 'S012b', 'SP2', 'SPAN', '1746525610000', '1746525650000', 'inst2', 1, 'P200', 'TASK200', 'INST200', 'ray'),
-- ===== T003: 正例 (positive example) — cross-day valid ray notebook-runner trace.
--   runner.execute: 05-06 19:00 -> 05-07 09:00 (orig); killed.end 05-07 10:00 forces
--     end_time correction (execute end -> 10:00) => observable instance metrics.
--   runner.killed: orig start 05-07 08:00, but code-group MIN start is 05-06 20:00
--     (execute.code) => killed.start correction pulls it back a day, moving the killed
--     span's day-split onto 05-06 and changing code_run_time/code_end on both days.
--   status_code from runner.execute only (=5; killed=7, execute.code=9, spc=0).
('2026050700', 'T003', 'SERVER', '0', 'notebook-runner', '', '', 'runner.execute', 3600, 'S030', 'SP3', 'SPAN', '1746529200000', '1746579600000', 'inst3', 5, 'P300', 'TASK300', 'INST300', 'ray'),
('2026050700', 'T003', 'SERVER', '0', 'notebook-runner', '', '', 'runner.killed', 7200, 'S031', 'SP3', 'SPAN', '1746576000000', '1746583200000', 'inst3', 7, 'P300', 'TASK300', 'INST300', 'ray'),
('2026050700', 'T003', 'SERVER', '0', 'notebook-runner', '', '', 'execute.code', 7200, 'S032', 'SP3', 'SPAN', '1746532800000', '1746540000000', 'inst3', 9, 'P300', 'TASK300', 'INST300', 'ray'),
('2026050700', 'T003', 'SERVER', '0', 'notebook-runner', '', '', 'set.permanent.compute', 300, 'S033', 'SP3', 'SPAN', '1746529500000', '1746529800000', 'inst3', 0, 'P300', 'TASK300', 'INST300', 'ray'),
-- ===== 诱饵 (decoy) — trace subquery filters. Each trace is excluded by exactly ONE
--   subquery condition (compute_type='ray' / service_name='notebook-runner' /
--   span_name='runner.execute'); if that condition is dropped, the trace leaks into output.
('2026050700', 'T004', 'SERVER', '0', 'notebook-runner', '', '', 'runner.execute', 1800, 'S040', 'SP4', 'SPAN', '1746590400000', '1746592200000', 'inst4', 1, 'P400', 'TASK400', 'INST400', 'cpu'),
('2026050700', 'T005', 'SERVER', '0', 'other-svc', '', '', 'runner.execute', 1800, 'S050', 'SP5', 'SPAN', '1746590400000', '1746592200000', 'inst5', 1, 'P500', 'TASK500', 'INST500', 'ray'),
('2026050700', 'T006', 'SERVER', '0', 'notebook-runner', '', '', 'execute.code', 1800, 'S060', 'SP6', 'SPAN', '1746590400000', '1746592200000', 'inst6', 1, 'P600', 'TASK600', 'INST600', 'ray'),
-- ===== 诱饵 (decoy) on included trace T001 — must be filtered by the outer WHERE.
--   rule: main databus_imp_date window [2026050400, 2026050700] — this row is 2026050800 (out of window).
('2026050800', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'execute.code', 300, 'S001w', 'SP1', 'SPAN', '1746756000000', '1746756300000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray'),
--   rule: span_name allowlist — 'runner.heartbeat' is NOT in the allowlist; lands on a fresh day (2025-05-08)
--   so if the allowlist filter is dropped it leaks an extra (T001, 2025-05-08) output row.
('2026050700', 'T001', 'SERVER', '0', 'notebook-runner', '', '', 'runner.heartbeat', 1800, 'S001h', 'SP1', 'SPAN', '1746684000000', '1746685800000', 'inst1', 0, 'P100', 'TASK100', 'INST100', 'ray');

-- ==================== Input table 2: notebook_engine_info ====================
CREATE TABLE IF NOT EXISTS notebook_engine_info_mysql_014 (
    `databus_imp_date` VARCHAR(256) COMMENT 'primary partition',
    `trace_id` VARCHAR(256) COMMENT 'trace_id',
    `service_name` VARCHAR(256) COMMENT 'service_name',
    `compute_type` VARCHAR(256) COMMENT 'compute_type',
    `datawd_project_id` VARCHAR(256) COMMENT 'datawd_project_id',
    `datawd_task_id` VARCHAR(256) COMMENT 'datawd_task_id',
    `datawd_task_instance_id` VARCHAR(256) COMMENT 'datawd_task_instance_id',
    `is_permanent` VARCHAR(256) COMMENT 'is_permanent',
    `serving_id` VARCHAR(256) COMMENT 'serving_id',
    `user_name` VARCHAR(256) COMMENT 'user_name',
    `is_model_service` VARCHAR(256) COMMENT 'is_model_service',
    `compute_info` VARCHAR(256) COMMENT 'compute_info',
    `num_gpu` VARCHAR(256) COMMENT 'num_gpu',
    `replicas` VARCHAR(256) COMMENT 'replicas',
    `gpu_type` VARCHAR(256) COMMENT 'gpu_type',
    `timestamp` VARCHAR(256) COMMENT 'timestamp',
    `ray_job_id` VARCHAR(256) COMMENT 'ray_job_id',
    `reason` VARCHAR(256) COMMENT 'reason',
    `log_url` VARCHAR(256) COMMENT 'log_url',
    `dashboard_url` VARCHAR(256) COMMENT 'dashboard_url'
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;

-- Mock data for notebook_engine_info
INSERT INTO notebook_engine_info_mysql_014 VALUES
('2026050700', 'T001', 'notebook-runner', 'ray', 'P100', 'TASK100', 'INST100', 'true', 'SRV001', 'user1', 'false', '{}', '2', '1', 'A100', '1746612000', 'JOB1', '', '', ''),
('2026050700', 'T002', 'notebook-runner', 'ray', 'P200', 'TASK200', 'INST200', 'false', 'SRV002', 'user2', 'false', '{}', '4', '2', 'V100', '1746525600', 'JOB2', '', '', ''),
-- T003: two valid engine rows so aggregations are non-trivial —
--   SUM(replicas*num_gpu)=3*2 + 2*4 = 14 (each single value 6/8 differs => covers SUM vs MIN/MAX),
--   MAX(serving_id)='SRV003B' (MIN='SRV003A' differs => covers serving_id MAX),
--   MAX(is_permanent)='true'  (MIN='false' differs => covers is_permanent MAX).
('2026050700', 'T003', 'notebook-runner', 'ray', 'P300', 'TASK300', 'INST300', 'false', 'SRV003A', 'user3', 'false', '{}', '2', '3', 'A100', '1746529200', 'JOB3', '', '', ''),
('2026050700', 'T003', 'notebook-runner', 'ray', 'P300', 'TASK300', 'INST300', 'true', 'SRV003B', 'user3', 'false', '{}', '4', '2', 'A100', '1746529200', 'JOB3', '', '', ''),
-- 诱饵 (decoy) engine row for T003 — compute_type='cpu' must be filtered out by the
--   engine WHERE (compute_type='ray' AND service_name='notebook-runner'). If that filter
--   is dropped, serving_id/is_permanent MAX and the gpu SUM all change (ZZZ999/zzz/+100).
('2026050700', 'T003', 'notebook-runner', 'cpu', 'P300', 'TASK300', 'INST300', 'zzz', 'ZZZ999', 'user3', 'false', '{}', '10', '10', 'A100', '1746529200', 'JOB3', '', '', '');

-- ==================== Output table: dwd_notebook_execute_instance_detail_d ====================
CREATE TABLE IF NOT EXISTS dwd_notebook_execute_instance_detail_d_mysql_014 (
    `dt` VARCHAR(256) COMMENT '天分区',
    `p_date` VARCHAR(256) COMMENT '拆分后日期',
    `trace_id` VARCHAR(256) COMMENT 'trace_id',
    `datawd_project_id` VARCHAR(256) COMMENT '项目ID',
    `datawd_task_id` VARCHAR(256) COMMENT '任务ID',
    `datawd_task_instance_id` VARCHAR(256) COMMENT '任务实例ID',
    `compute_type` VARCHAR(256) COMMENT '计算类型',
    `status_code` INT COMMENT '状态码',
    `instance_run_time` INT COMMENT '实例运行时长(秒)',
    `code_run_time` INT COMMENT '代码执行时长(秒)',
    `resource_wait_time` INT COMMENT '资源等待时长(秒)',
    `code_start_time` VARCHAR(256) COMMENT '代码开始时间',
    `code_end_time` VARCHAR(256) COMMENT '代码结束时间',
    `instance_start_time` VARCHAR(256) COMMENT '实例开始时间',
    `instance_end_time` VARCHAR(256) COMMENT '实例结束时间',
    `serving_id` VARCHAR(256) COMMENT 'serving_id',
    `is_permanent` VARCHAR(256) COMMENT '是否永久引擎',
    `apply_for_gpu_count` INT COMMENT 'GPU申请数量'
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;