-
Notifications
You must be signed in to change notification settings - Fork 2
/
demo-interactive-inspection.py
346 lines (277 loc) · 12.4 KB
/
demo-interactive-inspection.py
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
"""
Small demo for inspection of finished benchmarks.
Copyright (C) 2020 Patrick Erdelt
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as
published by the Free Software Foundation, either version 3 of the
License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
"""
from dbmsbenchmarker import *
import pandas as pd
#import logging
#logging.basicConfig(level=logging.DEBUG)
###################
##### general properties
###################
# path of folder containing experiment results
resultfolder = "./"
# create evaluation object for result folder
evaluate = inspector.inspector(resultfolder)
# list of all experiments in folder
evaluate.list_experiments
# dataframe of experiments
evaluate.get_experiments_preview()
# pick last experiment
code = evaluate.list_experiments[len(evaluate.list_experiments)-1]
# load it
evaluate.load_experiment(code)
###################
##### general experiment properties
###################
# hardware metrics
# list all available metrics
monitor.metrics.metrics
# get experiment workflow
df = evaluate.get_experiment_workflow()
# get workload properties
workload_properties = evaluate.get_experiment_workload_properties()
workload_properties['code']
workload_properties['path']
workload_properties['name']
workload_properties['info']
workload_properties['intro']
workload_properties['connectionmanagement']
workload_properties['reporting']
# list queries
list_queries = evaluate.get_experiment_list_queries()
# list connections
list_nodes = evaluate.get_experiment_list_nodes()
list_dbms = evaluate.get_experiment_list_dbms()
list_connections = evaluate.get_experiment_list_connections()
list_connections_node = evaluate.get_experiment_list_connections_by_node()
list_connections_dbms = evaluate.get_experiment_list_connections_by_dbms()
list_connections_clients = evaluate.get_experiment_list_connections_by_connectionmanagement('numProcesses')
list_connections_gpu = evaluate.get_experiment_list_connections_by_hostsystem('GPU')
# fix some examples:
# first connection, first query, first run
connection = list_connections[0]
numQuery = 1
numRun = 0
# show pretty some values
evaluator.pretty(evaluate.get_experiment_list_connections_by_dbms())
evaluator.pretty(evaluate.get_experiment_connection_properties(connection))
# get all dbms properties
# address: connection_properties[connection]['name']
connection_properties = evaluate.get_experiment_connection_properties()
# get single connection properties
connection_properties = evaluate.get_experiment_connection_properties(connection)
connection_properties['name']
connection_properties['hostsystem']
connection_properties['connectionmanagement']
connection_properties['connectionmanagement']['numProcesses']
# get single connection initialization script
evaluate.get_experiment_initscript(connection)
# get all queries properties
# address: query_properties[str(numQuery)]['title']
query_properties = evaluate.get_experiment_query_properties()
query_properties[str(numQuery)]['title']
# get single query properties
query_properties = evaluate.get_experiment_query_properties(numQuery)
query_properties['title']
query_properties['numRun']
query_properties['numWarmup']
query_properties['numCooldown']
query_properties['timer']
query_properties['query']
query_properties['parameter']
query_properties['active']
###################
##### general survey
###################
# some dataframes
# showing per DBMS and query overview
#
# measures, show as table
evaluate.get_aggregated_query_statistics(type='timer', name='run', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='throughput', name='run', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='latency', name='run', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='timer', name='connection', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='timer', name='execution', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='timer', name='datatransfer', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='monitoring', name='total_gpu_util', query_aggregate='Mean')
evaluate.get_aggregated_query_statistics(type='monitoring', name='total_gpu_memory', query_aggregate='Max')
# measures, show as heatmap (normalized)
evaluate.get_aggregated_query_statistics(type='timer', name='run', query_aggregate='factor')
# example: filter DBMS
list_omnisci = list_connections_dbms['OmniSci']
evaluate.get_aggregated_query_statistics(type='monitoring', name='total_gpu_memory', query_aggregate='factor', factor_base='Max', dbms_filter=list_omnisci)
# size of result sets
evaluate.get_total_resultsize()
# relative size of result sets (minimum=100%), show as heatmap
evaluate.get_total_resultsize_normalized()
# if there was an error, show as heatmap
evaluate.get_total_errors()
# if there was a warning, show as heatmap
evaluate.get_total_warnings()
# total times benchmarking a query took, show as table
evaluate.get_total_times()
# same, normalized to 100% per query, show as stacked area plot
evaluate.get_total_times_normalized()
# same, normed to relative to best per query, show as heatmap
evaluate.get_total_times_relative()
# get hardware metrics for loading test
evaluate.get_loading_metrics('total_cpu_memory')
# some dataframes
# showing metrics per dbms by aggregation
#
# from measures, show as bar plot
evaluate.get_aggregated_experiment_statistics(type='timer', name='run', query_aggregate='Mean', total_aggregate='Mean')
evaluate.get_aggregated_experiment_statistics(type='throughput', name='run', query_aggregate='Mean', total_aggregate='Geo')
evaluate.get_aggregated_experiment_statistics(type='timer', name='run', query_aggregate='factor', total_aggregate='Geo')
evaluate.get_aggregated_experiment_statistics(type='monitoring', name='total_gpu_memory', query_aggregate='Mean', total_aggregate='Mean')
# average data obtained from monitoring
# average per second - this differs from mean of mean, because not all queries have the same duration
evaluate.get_survey_monitoring()
# host data obtained from config
evaluate.get_survey_hostdata()
# example for merging two dataframes (by index)
df1 = evaluate.get_survey_monitoring()
df2 = evaluate.get_survey_hostdata()
tools.dataframehelper.merge(df1, df2)
###################
##### per query details
###################
# get measures of a query
#
# df1 = measures for plot, boxplot, histogramm
# df2 = statistics as table
df1, df2 = evaluate.get_measures_and_statistics(numQuery, type='timer', name='connection')
evaluate.get_measures_and_statistics(numQuery, type='timer', name='execution')
evaluate.get_measures_and_statistics(numQuery, type='timer', name='datatransfer')
evaluate.get_measures_and_statistics(numQuery, type='timer', name='run')
evaluate.get_measures_and_statistics(numQuery, type='timer', name='session')
evaluate.get_measures_and_statistics(numQuery, type='throughput', name='session')
evaluate.get_measures_and_statistics(numQuery, type='monitoring', name='total_gpu_util')
evaluate.get_measures_and_statistics(numQuery, type='monitoring', name='total_cpu_throttled')
evaluate.get_measures_and_statistics(numQuery, type='monitoring', name='total_gpu_memory')
evaluate.get_measures_and_statistics(numQuery, type='monitoring', name='total_gpu_memory', dbms_filter=list_omnisci, factor_base='Min')
# example: filter dbms, ignore 1 warmup
df1,df2=evaluate.get_measures_and_statistics(6, type='timer', name='run', warmup=1, dbms_filter=list_connections_dbms['MemSQL'])
# example: add statistics per column
evaluator.addStatistics(df2.T).T
# get errors for query
list_errors = evaluate.get_error(numQuery)
# restrict to connection
list_errors[connection]
# get warnings for query
list_warnings = evaluate.get_warning(numQuery)
# restrict to connection
list_warnings[connection]
# get data storage for query as list
list_storage = evaluate.get_datastorage_list(numQuery)
# restrict to run
list_storage[numRun]
# get data storage for query and run as dataframe
df = evaluate.get_datastorage_df(numQuery, numRun)
# get data storage for query and run as dataframe
# this only contains results if they differ from data storage
df2 = evaluate.get_resultset_df(numQuery, connection, numRun)
inspector.getDifference12(df2,df)
# get (sql) parameters of a query as dataframe
df = evaluate.get_parameter_df(numQuery)
# get size of data storage for query
evaluate.get_datastorage_size(numQuery)
# get size of received data of query in bytes
list_received_sizes = evaluate.get_received_sizes(numQuery)
# restrict to connection
list_received_sizes[connection]
# pretty print it
tools.sizeof_fmt(list_received_sizes[connection])
###################
##### per query aggregation
###################
# collect 3 aggregated metrics for query
df1, df2 = evaluate.get_measures_and_statistics(numQuery, type='timer', name='run')
df = tools.dataframehelper.collect(df2, 'Mean', 'timer_run_mean')
df1, df2 = evaluate.get_measures_and_statistics(numQuery, type='timer', name='session')
df = tools.dataframehelper.collect(df2, 'Mean', 'timer_session_mean', df)
df1, df2 = evaluate.get_measures_and_statistics(numQuery, type='throughput', name='session')
df = tools.dataframehelper.collect(df2, 'Median', 'throughput_session_median', df)
# show as table and bar chart
df
###################
##### examples
###################
# colors by dbms
list_connections_dbms = evaluate.get_experiment_list_connections_by_dbms()
connection_colors = evaluate.get_experiment_list_connection_colors(list_connections_dbms)
#connection_colors = evaluate.get_experiment_list_connection_colors(list_connections_node)
numQuery=21
df1,df2=evaluate.get_measures_and_statistics(numQuery, type='timer', name='run', warmup=0)
# using colors in matplotlib plots
#import matplotlib.pyplot as plt
#df1.T.plot(color=[connection_colors.get(x, '#333333') for x in df1.T.columns])
#plt.show()
# Some plotly figures
import plotly.graph_objects as go
# Plots
fig = go.Figure()
for i in range(len(df1.index)):
t = fig.add_trace(go.Scatter(x=df1.T.index, y=df1.iloc[i], name=df1.index[i], line=dict(color=connection_colors[df1.index[i]], width=1)))
fig.show()
# Boxplots
fig = go.Figure()
for i in range(len(df1.index)):
t = fig.add_trace(go.Box(y=df1.iloc[i], name=df1.index[i], line=dict(color=connection_colors[df1.index[i]], width=1), boxmean='sd'))
fig.show()
# Histograms
fig = go.Figure(layout = go.Layout(barmode='overlay'))
for i in range(len(df1.index)):
t = fig.add_trace(go.Histogram(x=df1.iloc[i], name=df1.index[i], opacity=0.75, marker=dict(color=connection_colors[df1.index[i]])))
fig.show()
# Table of statistics
df=df2.applymap(lambda x: ("%.2f" % float(x)))
df_2 = df.reset_index()
fig = go.Figure(data=[go.Table(
header=dict(values=list(df_2.columns),
fill_color='paleturquoise',
align='left'),
cells=dict(values=df_2.T.values.tolist(),
fill_color='lavender',
align='left'))
])
fig.show()
# Table of aggregated measures
df = evaluate.get_aggregated_query_statistics(type='timer', name='run', query_aggregate='Mean')
df=df.applymap(lambda x: ("%.2f" % x))
df_2 = df.reset_index()
fig = go.Figure(data=[go.Table(
header=dict(values=list(df_2.columns),
fill_color='paleturquoise',
align='left'),
cells=dict(values=df_2.T.values.tolist(),
fill_color='lavender',
align='left'))
])
fig.show()
# Heatmap
df = evaluate.get_aggregated_query_statistics(type='timer', name='run', query_aggregate='factor')
df = df.sort_index()
df_2=df.applymap(lambda x: ("%.2f" % x))
fig = go.Figure(data=[go.Heatmap(z=df_2.T.values.tolist(),x=df_2.index, y=df_2.columns,colorscale='Reds')])
# optionally fixed size
#t = fig.update_layout(autosize=False,height=2000)
fig.show()
# Bar
df1, df2 = evaluate.get_measures_and_statistics(numQuery, type='timer', name='run')
df = tools.dataframehelper.collect(df2, 'Mean', 'timer_run_mean')
fig = go.Figure()
for i in range(len(df.index)):
t = fig.add_trace(go.Bar(x=[df.index[i]], y=df.iloc[i], name=df.index[i], marker=dict(color=connection_colors[df.index[i]])))
fig.show()