Created
June 22, 2022 19:23
-
-
Save LouisJenkinsCS/8b3992f4c524a5f48ee0ea28ec5096b7 to your computer and use it in GitHub Desktop.
Convert JSON from mpstat command into a graph that shows CPU utilization over time (lightweight) while also giving maximum and minimum snapshots
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env python | |
# coding: utf-8 | |
import argparse | |
import pandas | |
import sys | |
import subprocess | |
import time | |
import json | |
import signal | |
import numpy as np | |
parser = argparse.ArgumentParser(description='Takes JSON from mpstat and creates and returns the minimum sample, maximum sample, and a 3D wireframe representation of the samples obtained. Used to characterize the data.') | |
parser.add_argument('input', metavar='i', type=str, help='The JSON file obtained from mpstat.') | |
parser.add_argument('--output', metavar='i', type=str, help="The output for the generated graph; Defaults to out.svg", default='out.svg') | |
args = parser.parse_args() | |
logfile = open(args.input, 'r+') | |
logfile.seek(0) | |
res = json.loads(logfile.read()) | |
data = res['sysstat']['hosts'][0]['statistics'] | |
df = pandas.DataFrame() | |
dfs = [] | |
for sample in data: | |
ts = sample['timestamp'] | |
loads = sample['cpu-load'] | |
dfs.append(pandas.DataFrame.from_dict(loads)) | |
cpu_load = {} | |
for df in dfs: | |
mapping = [] | |
for data in df['cpu']: | |
mapping.append(str(data)) | |
for col in df.columns: | |
if col == 'cpu': | |
continue | |
ix = 0 | |
for data in df[col]: | |
if mapping[ix] not in cpu_load: | |
cpu_load[mapping[ix]] = {} | |
if col not in cpu_load[mapping[ix]]: | |
cpu_load[mapping[ix]][col] = 0 | |
cpu_load[mapping[ix]][col] += data | |
ix += 1 | |
# Print out the average statistics | |
for x in cpu_load: | |
for y in cpu_load[x]: | |
cpu_load[x][y] /= len(dfs) | |
print(pandas.DataFrame.from_dict(cpu_load).transpose()) | |
# Print out the maximum and minimum dataframes | |
minUsr = 10000000 | |
maxUsr = 0 | |
maxDF = None | |
minDF = None | |
for df in dfs: | |
allUsr = df['usr'][0] | |
if maxUsr < allUsr: | |
maxDF = df | |
maxUsr = allUsr | |
if minUsr > allUsr: | |
minDF = df | |
minUsr = allUsr | |
print("Maximum Snapshot:\n", maxDF) | |
print("Minimum Snapshot:\n", minDF) | |
# Plot the data | |
import matplotlib.pyplot as plt | |
from mpl_toolkits.mplot3d import Axes3D | |
N=len(dfs[0]) | |
points = {k: [] for k in range(0, N)} | |
for df in dfs: | |
for col in range(0, N): | |
points[col].append(df['usr'][col]) | |
fig = plt.figure(figsize=(20,16)) | |
ax = Axes3D(fig, auto_add_to_figure=False) | |
fig.add_axes(ax) | |
ax.set_xlabel('Processor') | |
ax.set_ylabel('Seconds') | |
ax.set_zlabel('Utilization (%)') | |
xs = [] | |
ys = [] | |
zs = [] | |
for n in range(0,N): | |
tmp = points[n] | |
tmp = sorted(tmp) | |
_xs = [n for i in range(0, len(tmp))] | |
_ys = [i for i in range(0, len(tmp))] | |
_zs = tmp | |
d = np.array(list(zip(_xs,_ys,_zs))) | |
xs.append(d[:,0]) | |
ys.append(d[:,1]) | |
zs.append(np.array(d[:,2])) | |
xs = np.array(xs) | |
ys = np.array(ys) | |
zs = np.array(zs) | |
ax.plot_wireframe(xs, ys, zs, color='blue') | |
plt.savefig(args.output) | |
plt.show() | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment