-
Notifications
You must be signed in to change notification settings - Fork 6
/
DASH_Sim_utils.py
287 lines (265 loc) · 12.1 KB
/
DASH_Sim_utils.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
'''!
@brief This file contains functions that are used by DASH_Sim.
'''
import os
import csv
import fnmatch
import sys
import common
trace_list = [common.TRACE_FILE_SYSTEM, common.TRACE_FILE_TASKS, common.TRACE_FILE_FREQUENCY, common.TRACE_FILE_PES, common.TRACE_FILE_TEMPERATURE, common.TRACE_FILE_LOAD, common.TRACE_FILE_TEMPERATURE_WORKLOAD]
def update_PE_utilization_and_info(PE, current_timestamp):
'''!
Update the PE utilization.
@param PE: PE to be evaluated
@param current_timestamp: Current timestamp
'''
lower_bound = current_timestamp-common.sampling_rate # Find the lower bound for the time window_list under consideration
completed_info = []
running_info = []
for task in common.TaskQueues.completed.list:
#print('Time %s:'%current_timestamp, task.start_time, task.finish_time, task.PE_ID)
if task.PE_ID == PE.ID:
if ((task.start_time < lower_bound) and (task.finish_time < lower_bound)):
continue
elif ((task.start_time < lower_bound) and (task.finish_time >= lower_bound)):
completed_info.append(lower_bound)
completed_info.append(task.finish_time)
else:
completed_info.append(task.start_time)
completed_info.append(task.finish_time)
#print('Time %s:'%current_timestamp,'completed',completed_info, 'Task', task.ID, 'PE', PE.ID)
for task in common.TaskQueues.running.list:
#print('Time %s:'%current_timestamp, task.start_time, task.PE_ID)
if task.PE_ID == PE.ID:
if (task.start_time < lower_bound):
running_info.append(lower_bound)
else:
running_info.append(task.start_time)
task_start_time = task.start_time
running_info.append(current_timestamp)
#print('Time %s:'%current_timestamp,'running',running_info, 'Task', task.ID, 'PE', PE.ID)
merged_list = completed_info + running_info
# get the utilization for the PE
sum_of_active_times = sum([merged_list[i*2+1] - merged_list[i*2] for i in range(int(len(merged_list)/2))])
PE.utilization = (sum_of_active_times/common.sampling_rate) / PE.capacity
# print(PE.ID, PE.utilization)
full_list = [PE.ID, PE.utilization, current_timestamp]
info_list = [0 if i > (len(merged_list)-1) else merged_list[i] for i in range(12)]
full_list.extend(info_list)
PE.info = info_list
# if (common.DEBUG_SIM):
# print('Time %s: for PE-%d'%(current_timestamp,PE.ID),PE.info)
def trace_frequency(timestamp):
'''!
Trace method for saving the frequency variations.
@param timestamp: Current timestamp
'''
if (common.TRACE_FREQUENCY):
create_header = False
if not (os.path.exists(common.TRACE_FILE_FREQUENCY)):
# Create the CSV header
create_header = True
with open(common.TRACE_FILE_FREQUENCY, 'a', newline='') as csvfile:
trace = csv.writer(csvfile, delimiter=',')
if create_header == True:
header_list = ['Timestamp']
for idx, current_cluster in enumerate(common.ClusterManager.cluster_list):
if current_cluster.type != "MEM":
header_list.append('f_PE_' + str(idx))
header_list.append('N_PE_' + str(idx))
trace.writerow(header_list)
data = [timestamp]
for idx, current_cluster in enumerate(common.ClusterManager.cluster_list):
if current_cluster.type != "MEM":
data.append(current_cluster.current_frequency / 1000)
data.append(current_cluster.num_active_cores)
trace.writerow(data)
def trace_tasks(task, PE, task_time, total_energy):
'''!
Trace method for saving statistics about the tasks.
@param task: Task to be traced
@param PE: Current PE
@param task_time: Task's execution time
@param total_energy: Task's total energy consumption
'''
if (common.TRACE_TASKS):
create_header = False
if not (os.path.exists(common.TRACE_FILE_TASKS.split(".")[0] + "__" + str(common.trace_file_num) + ".csv")):
# Create the CSV header
create_header = True
with open(common.TRACE_FILE_TASKS.split(".")[0] + "__" + str(common.trace_file_num) + ".csv", 'a', newline='') as csvfile:
trace = csv.writer(csvfile, delimiter=',')
if create_header == True:
trace.writerow(['DVFS policy', 'Task ID', 'PE', 'Exec. Time (us)', 'Energy (J)'])
trace.writerow([common.ClusterManager.cluster_list[PE.cluster_ID].DVFS, task.ID, common.ClusterManager.cluster_list[PE.cluster_ID].name, task_time, total_energy])
def trace_system():
'''!
Trace method for saving statistics related to the system, i.e., the whole simulation.
'''
if (common.TRACE_SYSTEM):
create_header = False
if not (os.path.exists(common.TRACE_FILE_SYSTEM.split(".")[0] + "__" + str(common.trace_file_num) + ".csv")):
# Create the CSV header
create_header = True
with open(common.TRACE_FILE_SYSTEM.split(".")[0] + "__" + str(common.trace_file_num) + ".csv", 'a', newline='') as csvfile:
trace = csv.writer(csvfile, delimiter=',')
if create_header == True:
trace.writerow(['Job List', 'DVFS mode', 'N_little', 'N_big', 'Exec. Time (us)', 'Cumulative Exec. Time (us)', 'Energy (J)'])
DVFS_mode_list = []
for DVFS_config in common.DVFS_cfg_list:
if DVFS_config == "performance":
DVFS_mode_list.append("P")
elif DVFS_config == "powersave":
DVFS_mode_list.append("LP")
elif DVFS_config == "ondemand":
DVFS_mode_list.append("OD")
elif str(DVFS_config).startswith("constant"):
split = str(DVFS_config).split('-')
DVFS_mode_list.append("C" + split[1])
if common.simulation_mode == "validation":
trace.writerow([common.current_job_list, DVFS_mode_list, common.gen_trace_capacity_little, common.gen_trace_capacity_big,
common.results.execution_time, common.results.execution_time, common.results.energy_consumption])
elif common.simulation_mode == "performance":
if len(common.job_list) == 1:
job_list = common.current_job_list
else:
job_list = common.job_list
trace.writerow([job_list, DVFS_mode_list, common.gen_trace_capacity_little, common.gen_trace_capacity_big,
common.results.execution_time - common.warmup_period, common.results.cumulative_exe_time,
common.results.cumulative_energy_consumption])
def trace_PEs(timestamp, PE):
'''!
Trace method for saving statistics related to the PEs.
@param timestamp: Current timestamp
@param PE: PE to be traced
'''
if (common.TRACE_PES):
create_header = False
if not (os.path.exists(common.TRACE_FILE_PES)):
# Create the CSV header
create_header = True
with open(common.TRACE_FILE_PES, 'a', newline='') as csvfile:
dataset = csv.writer(csvfile, delimiter=',')
if create_header == True:
dataset.writerow(['Timestamp', 'PE', 'Info'])
dataset.writerow([timestamp, PE.ID, PE.info])
def trace_temperature(timestamp):
'''!
Trace method for saving the temperature variations.
@param timestamp: Current timestamp
'''
if (common.TRACE_TEMPERATURE):
create_header = False
if not (os.path.exists(common.TRACE_FILE_TEMPERATURE)):
# Create the CSV header
create_header = True
with open(common.TRACE_FILE_TEMPERATURE, 'a', newline='') as csvfile:
dataset = csv.writer(csvfile, delimiter=',')
if create_header == True:
dataset.writerow(['Timestamp', 'Snippet', 'Temperature', 'Throttling_state'])
dataset.writerow([timestamp, common.current_job_list, max(common.current_temperature_vector), common.throttling_state])
def trace_load(timestamp, PEs):
'''!
Trace method for saving the load variations
@param timestamp: Current timestamp
@param PEs: List of PEs
'''
if (common.TRACE_LOAD):
create_header = False
if not (os.path.exists(common.TRACE_FILE_LOAD)):
# Create the CSV header
create_header = True
with open(common.TRACE_FILE_LOAD, 'a', newline='') as csvfile:
dataset = csv.writer(csvfile, delimiter=',')
if create_header == True:
header_list = ['Timestamp', 'Snippet']
for idx, current_cluster in enumerate(common.ClusterManager.cluster_list):
if current_cluster.type != "MEM":
header_list.append('N_tasks_PE_' + str(idx))
header_list.append('N_tasks_total')
dataset.writerow(header_list)
data = [timestamp, common.current_job_list]
total_num_tasks = 0
for idx, current_cluster in enumerate(common.ClusterManager.cluster_list):
if current_cluster.type != "MEM":
num_tasks = get_num_tasks_being_executed(current_cluster, PEs)
data.append(num_tasks)
total_num_tasks += num_tasks
data.append(total_num_tasks)
dataset.writerow(data)
def get_current_job_list():
'''!
Get the current snippet.
@return Current snippet
'''
# Get the current job list based on the snippet ID while injecting jobs
if common.job_list != []:
return common.job_list[common.snippet_ID_exec]
else:
return common.job_list
def get_num_tasks_being_executed(cluster, PEs):
'''!
Get the number of tasks that are being executed.
@param cluster: Current cluster
@param PEs: List of PEs
@return Number of tasks being executed
'''
# Get the number of tasks currently being executed in the cluster
num_tasks = 0
for PE_ID in cluster.PE_list:
if not PEs[PE_ID].idle:
num_tasks += 1
return num_tasks
def get_cluster_utilization(cluster, PEs):
'''!
Get the cluster utilization.
@param cluster: Current cluster
@param PEs: List of PEs
@return Cluster utilization
'''
# Get the cluster utilization
utilization = 0
for PE_ID in cluster.PE_list:
utilization += PEs[PE_ID].utilization
return utilization / len(cluster.PE_list)
def clean_traces():
'''!
Remove old trace files.
'''
for trace_name in trace_list:
if os.path.exists(trace_name):
os.remove(trace_name)
# Remove old traces generated in parallel
for trace_name in trace_list:
file_list = fnmatch.filter(os.listdir('.'), trace_name.split(".")[0] + '__*.csv')
for f in file_list:
os.remove(f)
def clean_policies():
'''!
Remove old policy files.
'''
file_list = fnmatch.filter(os.listdir('.'), '*.pkl')
for f in file_list:
os.remove(f)
def init_variables_at_sim_start() :
'''!
Initialize config variables.
'''
common.snippet_start_time = common.warmup_period
common.snippet_ID_inj = -1
common.snippet_ID_exec = 0
common.snippet_throttle = -1
common.snippet_temp_list = []
common.snippet_initial_temp = [common.T_ambient,
common.T_ambient,
common.T_ambient,
common.T_ambient,
common.T_ambient]
common.current_temperature_vector = [common.T_ambient, # Indicate the current PE temperature for each hotspot
common.T_ambient,
common.T_ambient,
common.T_ambient,
common.T_ambient]
common.B_model = []
common.job_counter_list = [0]*len(common.current_job_list)
common.throttling_state = -1