Author: Vitali Kremez
This iPython notebook creates and explores a SQL database of all known public exploits. Source: https://raw.githubusercontent.com/offensive-security/exploit-database/master/files.csv
PHP has the largest number of exploits in the SQL database. Function: (1) Creates and queries the SQL database with approx. 40,000 rows and 9 columns of public exploits; (2) Creates a graph "Number of remote exploits by platform"; [1] Windows OS has 3,340 remote exploits, the largest number of remote exploits in the SQL database. [2] Linux OS has 825 remote exploits, the second largest number in the database (3) Creates a graph "Number of webapps exploits by platform"; [1] We have approx. 17,335 of PHP webapp exploits, the largest number of webapp exploits in the SQL database. [2] We have only 1,498 asp webapp exploits, the second largest number in the database. (4) Creates a graph "Number of local exploits by platform"; [1] We have 1,561 of local Windows exploits, the largest number of local exploits in the SQL database. [2] We have only 819 Linux local exploits, the second largest number in the database. (5) Creates a graph "Number of Denial- of-Service (DOS) exploits by platform"; [1] We have 2,845 of DOS Windows exploits, the largest number of DOS exploits in the SQL database. [2] We have only 603 Linux DOS exploits, the second largest number in the database. (6) Creates a graph "Number of exploits by platform"; [1] We have approx. 17,590 of PHP exploits, the largest number of exploits in the SQL database. [2] We have only 603 Linux exploits, the second largest number in the database. (7) Queries the SQL database for Windows exploits; We have 7,921 Windows exploits in the database. (8) Creates a graph "Number of exploits by date"; We have the largest number of exploits developed in 2010 among other years. Windows has the largest number of DOS exploits among other platforms. 
In [39]:
import pandas as pd
from sqlalchemy import create_engine # database connection
import datetime as dt
from IPython.display import display
In [40]:
import plotly.plotly as py # interactive graphing
from plotly.graph_objs import Bar, Scatter, Marker, Layout
In [13]:
disk_engine = create_engine('sqlite:///ExploitDB.db')
In [14]:
start = dt.datetime.now()
chunksize = 40000
j = 0
index_start = 1
In [15]:
for df in pd.read_csv('https://raw.githubusercontent.com/offensive-security/exploit-database/master/files.csv', chunksize=chunksize, iterator=True, encoding='utf-8'):
df.index += index_start
j+=1
print '{} seconds: completed {} rows'.format((dt.datetime.now() - start).seconds, j*chunksize)
df.to_sql('data', disk_engine, if_exists='append')
index_start = df.index[-1] + 1
In [42]:
df = pd.read_sql_query('SELECT * FROM data', disk_engine)
print df
In [ ]:
In [18]:
df = pd.read_sql_query("SELECT platform, COUNT(*) as 'num_remote' FROM data WHERE type LIKE '%remote%' GROUP BY platform ORDER BY 'num_remote'", disk_engine)
py.iplot([Bar(x=df.platform, y=df.num_remote)], filename='Number of remote exploits by platform')
Out[18]:
In [20]:
df = pd.read_sql_query("SELECT platform, COUNT(*) as 'num_webapps' FROM data WHERE type LIKE '%webapps%' GROUP BY platform ORDER BY 'num_webapps'", disk_engine)
py.iplot([Bar(x=df.platform, y=df.num_webapps)], filename='Number of webapps exploits by platform')
Out[20]:
In [21]:
df = pd.read_sql_query("SELECT platform, COUNT(*) as 'num_local' FROM data WHERE type LIKE '%local%' GROUP BY platform ORDER BY 'num_local'", disk_engine)
py.iplot([Bar(x=df.platform, y=df.num_local)], filename='Number of local exploits by platform')
Out[21]:
In [22]:
df = pd.read_sql_query("SELECT platform, COUNT(*) as 'num_dos' FROM data WHERE type LIKE '%dos%' GROUP BY platform ORDER BY 'num_dos'", disk_engine)
py.iplot([Bar(x=df.platform, y=df.num_dos)], filename='Number of dos exploits by platform')
Out[22]:
In [24]:
df = pd.read_sql_query("SELECT platform, COUNT(*) as 'num_exploits' FROM data GROUP BY platform ORDER BY 'num_exploits'", disk_engine)
py.iplot([Bar(x=df.platform, y=df.num_exploits)], filename='Number of exploits by platform')
Out[24]:
In [35]:
df = pd.read_sql_query("SELECT platform, COUNT(*) as 'num_exploits' FROM data GROUP BY platform ORDER BY 'num_exploits'", disk_engine)
In [36]:
print df
In [16]:
df = pd.read_sql_query("SELECT file, COUNT(*) as 'num_windows' FROM data WHERE file LIKE '%windows%' GROUP BY file ORDER BY 'num_windows'", disk_engine)
In [17]:
print df
In [38]:
df = pd.read_sql_query("SELECT date, COUNT(*) as 'num_exploit' FROM data GROUP BY date ORDER BY 'num_exploit'", disk_engine)
py.iplot([Bar(x=df.date, y=df.num_exploit)], filename='Number of exploits by date')
Out[38]:
0 Comments
Leave a Reply. |
AuthorVitali Kremez Archives
January 2016
Categories |
RSS Feed
