Skip to content

Commit

Permalink
Merge pull request #69 from cracraft/filesystem-monitor
Browse files Browse the repository at this point in the history
Filesystem monitor
  • Loading branch information
bourque committed May 31, 2018
2 parents 8d9740b + 2e9d34c commit 9a3edea
Show file tree
Hide file tree
Showing 2 changed files with 248 additions and 0 deletions.
Empty file.
248 changes: 248 additions & 0 deletions jwql/monitor_filesystem/monitor_filesystem.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,248 @@
#! /usr/bin/env python

"""
This module is meant to monitor and gather statistics of the filesystem
that hosts data for the ``jwql`` application. This will answer
questions such as the total number of files, how much disk space is
being used, and then plot these values over time.
Authors
-------
- Misty Cracraft
Use
---
This module can be executed from the command line:
::
python monitor_filesystem.py
Alternatively, it can be called from scripts with the following
import statements:
::
from monitor_filesystem import filesystem_monitor
from monitor_filesystem import plot_system_stats
Required arguments (in a ``config.json`` file):
``filepath`` - The path to the input file needs to be in a
``config.json`` file in the ``utils`` directory
``outputs`` - The path to the output files needs to be in a
``config.json`` file in the ``utils`` directory.
Required arguments for plotting:
``inputfile`` - The name of the file to save all of the system
statistics to
``filebytype`` - The name of the file to save stats on fits type
files to
Dependencies
------------
The user must have a configuration file named ``config.json``
placed in the ``utils`` directory.
Notes
-----
The ``filesystem_monitor`` function queries the filesystem,
calculates the statistics and saves the output file(s) in the
directory specified in the ``config.json`` file.
The ``plot_system_stats`` function reads in the two specified files
of statistics and plots the figures to an html output page as well
as saving them to an output html file.
"""

from collections import defaultdict
import datetime
import numpy as np
import os
import subprocess

from bokeh.plotting import figure, output_file, save
from bokeh.layouts import gridplot

from jwql.permissions.permissions import set_permissions
from jwql.utils.utils import filename_parser
from jwql.utils.utils import get_config


def filesystem_monitor():
""" Get statistics on filesystem"""

# Get path, directories and files in system and count files in all directories
settings = get_config()
filesystem = settings['filesystem']
outputs_dir = os.path.join(settings['outputs'], 'filesystem_monitor')

# set up dictionaries for output
results_dict = defaultdict(int)
size_dict = defaultdict(float)
# Walk through all directories recursively and count files
for dirpath, dirs, files in os.walk(filesystem):
results_dict['file_count'] += len(files) # find number of all files
for filename in files:
file_path = os.path.join(dirpath, filename)
if filename.endswith(".fits"): # find total number of fits files
results_dict['fits_files'] += 1
size_dict['size_fits'] += os.path.getsize(file_path)
suffix = filename_parser(filename)['suffix']
results_dict[suffix] += 1
size_dict[suffix] += os.path.getsize(file_path)

# Get df style stats on file system
out = subprocess.check_output('df {}'.format(filesystem), shell=True)
outstring = out.decode("utf-8") # put into string for parsing from byte format
parsed = outstring.split(sep=None)

# Select desired elements from parsed string
total = int(parsed[8]) # in blocks of 512 bytes
used = int(parsed[9])
available = int(parsed[10])
percent_used = parsed[11]

# Save stats for plotting over time
now = datetime.datetime.now().isoformat(sep='T', timespec='auto') # get date of stats

# set up output file and write stats
statsfile = os.path.join(outputs_dir, 'statsfile.txt')
with open(statsfile, "a+") as f:
f.write("{0} {1:15d} {2:15d} {3:15d} {4:15d} {5}\n".format(now, results_dict['file_count'],
total, available, used, percent_used))
set_permissions(statsfile, verbose=False)

# set up and read out stats on files by type
filesbytype = os.path.join(outputs_dir, 'filesbytype.txt')
with open(filesbytype, "a+") as f2:
f2.write("{0} {1} {2} {3} {4} {5}\n".format(results_dict['fits_files'],
results_dict['uncal'], results_dict['cal'], results_dict['rate'],
results_dict['rateints'], results_dict['i2d']))
set_permissions(filesbytype, verbose=False)

# set up file size by type file
sizebytype = os.path.join(outputs_dir, 'sizebytype.txt')
with open(sizebytype, "a+") as f3:
f3.write("{0} {1} {2} {3} {4} {5}\n".format(size_dict['size_fits'],
size_dict['uncal'], size_dict['cal'], size_dict['rate'],
size_dict['rateints'], size_dict['i2d']))
set_permissions(sizebytype, verbose=False)


def plot_system_stats(stats_file, filebytype, sizebytype):
"""Read in the file of saved stats over time and plot them.
Parameters
-----------
stats_file : str
file containing information of stats over time
filebytype : str
file containing information of file counts by type over
time
sizebytype : str
file containing information on file sizes by type over time
"""

# get path for files
settings = get_config()
outputs_dir = os.path.join(settings['outputs'], 'filesystem_monitor')

# read in file of statistics
date, f_count, sysize, frsize, used, percent = np.loadtxt(os.path.join(outputs_dir, stats_file), dtype=str, unpack=True)
fits_files, uncalfiles, calfiles, ratefiles, rateintsfiles, i2dfiles = np.loadtxt(os.path.join(outputs_dir, filebytype), dtype=str, unpack=True)
fits_sz, uncal_sz, cal_sz, rate_sz, rateints_sz, i2d_sz = np.loadtxt(os.path.join(outputs_dir, sizebytype), dtype=str, unpack=True)

# put in proper np array types and convert to GB sizes
dates = np.array(date, dtype='datetime64')
file_count = f_count.astype(float)
systemsize = sysize.astype(float) / (1024.**3)
freesize = frsize.astype(float) / (1024.**3)
usedsize = used.astype(float) / (1024.**3)

fits = fits_files.astype(int)
uncal = uncalfiles.astype(int)
cal = calfiles.astype(int)
rate = ratefiles.astype(int)
rateints = rateintsfiles.astype(int)
i2d = i2dfiles.astype(int)

fits_size = fits_sz.astype(float) / (1024.**3)
uncal_size = uncal_sz.astype(float) / (1024.**3)
cal_size = cal_sz.astype(float) / (1024.**3)
rate_size = rate_sz.astype(float) / (1024.**3)
rateints_size = rateints_sz.astype(float) / (1024.**3)
i2d_size = i2d_sz.astype(float) / (1024.**3)

# plot the data
# Plot filecount vs. date
p1 = figure(
tools='pan,box_zoom,reset,save', x_axis_type='datetime',
title="Total File Counts", x_axis_label='Date', y_axis_label='Count')
p1.line(dates, file_count, line_width=2, line_color='blue')
p1.circle(dates, file_count, color='blue')

# Plot system stats vs. date
p2 = figure(
tools='pan,box_zoom,reset,save', x_axis_type='datetime',
title='System stats', x_axis_label='Date', y_axis_label='GB')
p2.line(dates, systemsize, legend='Total size', line_color='red')
p2.circle(dates, systemsize, color='red')
p2.line(dates, freesize, legend='Free bytes', line_color='blue')
p2.circle(dates, freesize, color='blue')
p2.line(dates, usedsize, legend='Used bytes', line_color='green')
p2.circle(dates, usedsize, color='green')

# Plot fits files by type vs. date
p3 = figure(
tools='pan,box_zoom,reset,save', x_axis_type='datetime',
title="Total File Counts by Type", x_axis_label='Date', y_axis_label='Count')
p3.line(dates, fits, legend='Total fits files', line_color='black')
p3.circle(dates, fits, color='black')
p3.line(dates, uncal, legend='uncalibrated fits files', line_color='red')
p3.diamond(dates, uncal, color='red')
p3.line(dates, cal, legend='calibrated fits files', line_color='blue')
p3.square(date, cal, color='blue')
p3.line(dates, rate, legend='rate fits files', line_color='green')
p3.triangle(dates, rate, color='green')
p3.line(dates, rateints, legend='rateints fits files', line_color='orange')
p3.asterisk(dates, rateints, color='orange')
p3.line(dates, i2d, legend='i2d fits files', line_color='purple')
p3.x(dates, i2d, color='purple')

# plot size of total fits files by type
p4 = figure(
tools='pan,box_zoom,reset,save', x_axis_type='datetime',
title="Total File Sizes by Type", x_axis_label='Date', y_axis_label='GB')
p4.line(dates, fits_size, legend='Total fits files', line_color='black')
p4.circle(dates, fits_size, color='black')
p4.line(dates, uncal_size, legend='uncalibrated fits files', line_color='red')
p4.diamond(dates, uncal_size, color='red')
p4.line(dates, cal_size, legend='calibrated fits files', line_color='blue')
p4.square(date, cal_size, color='blue')
p4.line(dates, rate_size, legend='rate fits files', line_color='green')
p4.triangle(dates, rate_size, color='green')
p4.line(dates, rateints_size, legend='rateints fits files', line_color='orange')
p4.asterisk(dates, rateints_size, color='orange')
p4.line(dates, i2d_size, legend='i2d fits files', line_color='purple')
p4.x(dates, i2d_size, color='purple')

# create a layout with a grid pattern
grid = gridplot([[p1, p2], [p3, p4]])
outfile = os.path.join(outputs_dir, "filesystem_monitor.html")
output_file(outfile)
save(grid)
set_permissions(outfile, verbose=False)


if __name__ == '__main__':

inputfile = 'statsfile.txt'
filebytype = 'filesbytype.txt'
sizebytype = 'sizebytype.txt'

filesystem_monitor()
plot_system_stats(inputfile, filebytype, sizebytype)

0 comments on commit 9a3edea

Please sign in to comment.