-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathmetrics_fetch.py
169 lines (146 loc) · 7.67 KB
/
metrics_fetch.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
#coding=utf-8
import urllib.parse
import requests
import os
import datetime
import time
import pandas as pd
import numpy as np
template = {
"vCPU": "sum(rate(container_cpu_usage_seconds_total{{namespace='{0}',container='{1}'}}[{2}]))",
"cpu":"sum(irate(container_cpu_usage_seconds_total{{container=~'{1}',namespace=~'{0}'}}[1m]))/sum(container_spec_cpu_quota{{container=~'{1}',namespace=~'{0}'}}/container_spec_cpu_period{{container=~'{1}',namespace=~'{0}'}})",
"mem": "sum(container_memory_usage_bytes{{namespace='{0}',container='{1}'}}) / sum(container_spec_memory_limit_bytes{{namespace='{0}',container='{1}'}})",
"mem_": "sum(container_memory_usage_bytes{{namespace='{0}',container='{1}'}})",
"res": "sum(rate(istio_request_duration_milliseconds_sum{{reporter='destination',destination_workload_namespace='{0}',destination_workload='{1}'}}[{2}]))/sum(rate(istio_request_duration_milliseconds_count{{reporter='destination',destination_workload_namespace='{0}',destination_workload='{1}'}}[{2}]))/1000",
"req": "sum(rate(istio_requests_total{{destination_workload_namespace='{0}',destination_workload='{1}'}}[{2}]))",
"pod": "count(container_spec_cpu_period{{namespace='{0}',container='{1}'}})"
}
prefix_api = "http://localhost:30090/api/v1/query?query="
namespace = 'boutiquessj'
interval = 120
services = ["adservice", "cartservice", "checkoutservice","currencyservice","emailservice","frontend","paymentservice","productcatalogservice","recommendationservice","shippingservice"]
metrics = ['cpu','res','req','pod']
training_root_dir = ''
def fetch_cpu_usage(svc_name, namespace=namespace, interval=30):
cpu_api = template["cpu"].format(namespace, svc_name, str(interval)+'s')
url = prefix_api + urllib.parse.quote_plus(cpu_api)
res = requests.get(url).json()["data"]
v = 0
if "result" in res and len(res["result"]) > 0 and "value" in res["result"][0]:
v = res["result"][0]["value"][1]
return float(v)
def fetch_mem_usage(svc_name, namespace=namespace):
mem_api = template["mem"].format(namespace, svc_name)
url = prefix_api + urllib.parse.quote_plus(mem_api)
res = requests.get(url).json()["data"]
v = 0
if "result" in res and len(res["result"]) > 0 and "value" in res["result"][0]:
v = res["result"][0]["value"][1]
return float(v)
def fetch_res_time(svc_name, namespace=namespace, interval=30):
res_api = template["res"].format(namespace, svc_name, str(interval)+'s')
url = prefix_api + urllib.parse.quote_plus(res_api)
res = requests.get(url).json()["data"]
if "result" in res and len(res["result"]) > 0 and "value" in res["result"][0]:
v = res["result"][0]["value"]
if v[1] != 'NaN':
return float(v[1])
return 0
def fetch_req(svc_name, namespace=namespace, interval=30):
req_api = template["req"].format(namespace, svc_name, str(interval)+'s')
url = prefix_api + urllib.parse.quote_plus(req_api)
req = requests.get(url).json()["data"]
if "result" in req and len(req["result"]) > 0 and "value" in req["result"][0]:
v = req["result"][0]["value"]
if v[1] != 'NaN':
return int(float(v[1]))
return 0
def fetch_prior_req(svc_name, namespace=namespace, interval=30, delta=30):
req_api = template["req"].format(namespace, svc_name, str(interval)+'s')
url = prefix_api + urllib.parse.quote_plus(req_api) + "&time=" + str(time.time() - delta)
req = requests.get(url).json()["data"]
if "result" in req and len(req["result"]) > 0 and "value" in req["result"][0]:
v = req["result"][0]["value"]
if v[1] != 'NaN':
return int(float(v[1]))
return 0
def fetch_pods(svc_name, namespace=namespace):
pod_api = template["pod"].format(namespace, svc_name)
url = prefix_api + urllib.parse.quote_plus(pod_api)
print(url)
#pod = requests.get(url).json()
pod = requests.get(url).json()["data"]
if "result" in pod and len(pod["result"]) > 0 and "value" in pod["result"][0]:
v = pod["result"][0]["value"]
if v[1] != 'NaN':
return int(float(v[1]))
return 0
def save_fetch_data(svc_name, mode, start_time, latsted_time, interval, save_file):
api_str = template[mode].format(namespace, svc_name, str(interval)+'s')#mode为pod，
# start_time = datetime.datetime.strptime(start_time, '%Y-%m-%d %H:%M:%S')
with open(save_file, 'w') as f:
for i in range(0, latsted_time, int(interval)):#以interval为间隔
t = start_time + datetime.timedelta(seconds=i)#加时间
unixtime = time.mktime(t.timetuple())#返回用秒数来表示时间的浮点数。
url = prefix_api + urllib.parse.quote_plus(api_str) + "&time=" + str(unixtime)
res = requests.get(url).json()["data"]
if "result" in res and len(res["result"]) > 0 and "value" in res["result"][0]:
v = res["result"][0]["value"]
if v[1] == 'NaN':
print("0", file=f)
else:
print(str(v[1]), file=f)
else:
print("0", file=f)
def save_all_fetch_data(times=[], start_iter=1, root_dir='/home/boutiquessj/pythonForK6/data/', interval=interval, services=services, metrics=metrics):
if not os.path.exists(root_dir):#若不存在则创建一个
os.makedirs(root_dir)
for i, (start_time, end_or_lasted) in enumerate(times):
start_time = datetime.datetime.strptime(start_time, '%Y-%m-%d %H:%M:%S') #+ datetime.timedelta(hours=8)
if type(end_or_lasted) == str:
end_time = datetime.datetime.strptime(end_or_lasted, '%Y-%m-%d %H:%M:%S') #+ datetime.timedelta(hours=8)#######因为时区问题要加8
lasted_time = int((end_time-start_time).total_seconds())#有多少秒
else:
lasted_time = int(end_or_lasted)
if os.path.exists(root_dir+'{}_productpage_res.log'.format(start_iter+i)):
print("Iternum needs update")
return
for svc in services:
for m in metrics:
save_fetch_data(svc, m, start_time, lasted_time, interval, root_dir+"{}_{}_{}.log".format(start_iter+i, svc, m))
print("saved file:"+root_dir+"{}_{}_{}.log".format(start_iter+i, svc, m))
def load_fetch_data(root_dir, start_iter=1, end_iter=None, services=services, metrics=metrics) -> pd.DataFrame:
if not end_iter:
end_iter = start_iter
data = {}
for svc in services:
data[svc] = {}
for m in metrics:
data[svc][m] = []
for svc in services:
for m in metrics:
for iternum in range(start_iter, end_iter+1):
path = root_dir + '{}_{}_{}.log'.format(iternum, svc, m)
with open(path, 'r') as f:
lines = f.readlines()
data[svc][m] += list(map(lambda x:float(x), lines))
D = [data[svc][m] for svc in services for m in metrics]
data_df = pd.DataFrame(np.array(D).T, columns=[svc+"_"+m for svc in services for m in metrics])
return data_df
def load_processed_fetch_data(iternums=[1, 2], root_dir=training_root_dir, metrics=metrics):
data_df = load_fetch_data(iternums, root_dir, metrics)
D, l = [], len(metrics)
for r in data_df.values:
D.append([1]+list(r[:l]))
D.append([2]+list(r[l:2*l]))
D.append([3]+list(r[2*l:3*l]))
D.append([4]+list(r[3*l:]))
data_df = pd.DataFrame(D, columns=['svc']+metrics)
data_df['pod'] = data_df['pod'].astype(int)
return data_df
if __name__ == '__main__':
times = [
('2023-03-04 04:09:01', '2023-03-04 05:25:15')
]#
save_all_fetch_data(times, 1, root_dir='/ssj/ssj/boutiquessj/pyboutique/newData/slohpa/', interval=30, services=services)#interval 间隔
print("ok")