blob: efa8320eae38b7f6a7bb5bf30e8de62afd937ff1 (
plain)
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
|
import threading
import time
import psutil
import sys
from .util import timestamp_now, get_build_metric_dict
from .protocol import BuildMetricsCollector
class MemoryMonitor(BuildMetricsCollector):
INTERVAL = 0.1 # seconds
def __init__(self, starting_memory_adjustment=0):
self._stop = False
metrics = get_build_metric_dict()
metrics['system_memory'] = {
"mem_over_time": [],
"start_mem": used_memory() - starting_memory_adjustment,
}
self._thread = threading.Thread(target=self.memory_monitor, daemon=True)
self._thread.start()
def finalize(self):
self._stop = True
self._record_data_point()
metrics = get_build_metric_dict()
sys_mem = metrics["system_memory"]
mean = 0
max_ = 0
count = 1
for val in sys_mem["mem_over_time"]:
max_ = max(val["memory"], max_)
# iterative mean calculation algorithm from https://stackoverflow.com/a/1934266
mean += (val["memory"] - mean) / count
count += 1
sys_mem["arithmetic_mean"] = mean
sys_mem["max"] = max_
def memory_monitor(self):
while not self._stop:
time.sleep(self.INTERVAL)
if self._stop:
break
self._record_data_point()
def _record_data_point(self):
used_mem = used_memory()
now_time = timestamp_now()
metrics = get_build_metric_dict()
metrics["system_memory"]["mem_over_time"].append(
{"timestamp": now_time, "memory": used_mem})
def used_memory():
return psutil.virtual_memory().used
|
