benchmarks/load.py at master · DITAS-Project/benchmarks · GitHub

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
#! /usr/bin/env python3

import pandas as pd
import numpy as np

import glob
from datetime import datetime
from dateutil.parser import parse

from elasticsearch import Elasticsearch

def _load_vmstat(monitoring_data):

    monitoring_data["timestamp"] = pd.to_datetime(monitoring_data["timestamp"]+ 3600, unit='s')
    monitoring_data = monitoring_data.rename(columns={"r":"processes","b":"waiting","swdp":"virtual mem","free":"free","buff":"buffers","si":"mem_on_disk","so":"mem_to_disk","bi":"blockIn","bo":"blockOut","in":"interrupts","cs":"switches","us":"cpu_user","sy":"cpu_system","id":"cpu_idle","wa":"blocked"})

    return monitoring_data

def load_vmstat(load_from_cache=False,store_cache_file=False,cache_file=None):
    monitoring_data = None

    if load_from_cache and cache_file is not None:
        monitoring_data = pd.read_csv(cache_file)
    else:

        for file in glob.glob("vmstats/*"):
            df = pd.read_csv(file, skiprows = 0,error_bad_lines=False)
            if monitoring_data is None:
                monitoring_data = df
            else:
                monitoring_data = pd.concat([monitoring_data, df], sort=True)
        #clean up data
        monitoring_data = _load_vmstat(monitoring_data)
        if store_cache_file:
            monitoring_data.to_csv(cache_file)

    return monitoring_data

def load_elastic(load_from_cache=False,store_cache_file=False,cache_file=None,es=None,experiment_dates=[]):
    monitoring_data = None

    if load_from_cache and cache_file is not None:
        monitoring_data = pd.read_csv(cache_file)
    else:
        monitoring_data = collect_monitoring_data(es,"*",experiment_dates)
        if store_cache_file:
            if monitoring_data is not None:
                monitoring_data.to_csv(cache_file)

    return monitoring_data

def load_rmstats():
    monitoring_data = None

    for file in glob.glob("rmstats/*.csv"):
            df = pd.read_csv(file, skiprows = 0,error_bad_lines=False)
            if monitoring_data is None:
                monitoring_data = df
            else:
                monitoring_data = pd.concat([monitoring_data, df], sort=True)

    return monitoring_data

def load_experiment(load_from_cache=False,store_cache_file=False,data_cache_file=None):
    data = None

    if load_from_cache and data_cache_file is not None:
        data = pd.read_csv(data_cache_file)
    else:
        data = __load()
        if store_cache_file:
            data.to_csv(data_cache_file)


    return data

def __load():
    all = None
    for file in glob.glob("data/*"):
        names = file[5:-4].split("-")
        experiment=names[0]
        method=names[1]
        datetime.strptime
        timestamp=file[5+len(experiment)+1+len(method)+1:-4]
        date=timestamp[:timestamp.index("T")]
        date=datetime.strptime(date, '%Y-%m-%d')
        timestamp=parse(timestamp)
        df = pd.read_csv(file, skiprows = 0,error_bad_lines=False)
        df['experiment']=experiment
        df['method']=method
        df['startTime']=pd.Timestamp(year=timestamp.year,month=timestamp.month, day=timestamp.day, hour=timestamp.hour, minute=timestamp.minute)
        df['runDate']=pd.Timestamp(year=date.year,month=date.month, day=date.day)

        if (all is None):
            all = df
        else:
            all = pd.concat([all, df], sort=True)

    return all

def __collect(es,index,query):
    data = []
    page = es.search(
    index = index,
    scroll = '2m',
    size = 1000,
    body = query)
    if '_scroll_id' in page:

        sid = page['_scroll_id']
        scroll_size = page['hits']['total']

        data = data + page['hits']['hits']
        # Start scrolling
        while (scroll_size > 0):
            page = es.scroll(scroll_id = sid, scroll = '2m')
            # Update the scroll ID
            sid = page['_scroll_id']
            # Get the number of results that we returned in the last scroll
            scroll_size = len(page['hits']['hits'])
            data = data + page['hits']['hits']

        return data
    else:
        return data

def collect_monitoring_data(es,vdcname,dates=[]):
    all = None
    for runDate in dates:
        esAll = []
        index = "{}-{}".format(vdcname,runDate.date().strftime("%Y-%m-%d"))
        print("loading data from index",index)

        esAll =  __collect(es,index,{"query": {"match_all": {}}})

        if len(esAll) <= 0:
            continue


        esAll = list(map(lambda x:x["_source"],esAll))


        responses = filter(lambda x:'response.code' in x,esAll)
        requests = filter(lambda x:'response.code' not in x,esAll)

        responses = pd.DataFrame(responses)
        responses = responses[['request.id','response.code','response.length']]
        requests = pd.DataFrame(requests)
        requests = requests[['request.id','@timestamp','request.operationID','request.requestTime','request.client','request.method','request.path']]
        df = pd.merge(requests, responses, on='request.id')
        df['runDate'] = runDate

        if (all is None):
            all = df
        else:
            all = pd.concat([all, df], sort=True)

    return all