The scheduling format for the Linux scheduling daemon cron are not easy to remember, especially if you don’t work with cron that frequently. The first reaction for most Linux sys admins when they can’t remember the ordering of fields is to type ‘man crontab’, and unfortunately this man page section does not contain the schedule format information. If you are like me, you will immediately start Googling it.

What is the best way to locate the man page for crontab scheduling format then? For one thing, you can search the man page for the key work ‘crontab’ using the command below –

daniel@linubuvma:/tmp$ man -k crontab
anacrontab (5)       - configuration file for anacron
crontab (1)          - maintain crontab files for individual users (Vixie Cron)
crontab (5)          - tables for driving cron

You see, there are two sections for crontab – section 1 describes the command usage and section 5 shows the tables we are looking for. If you are familiar with how man page section numbers are assigned, you would have immediately jumped to section 5 of the man page for crontab –


1. General commands
2. System calls
3. C library functions
4. Special files (usually devices, those found in /dev) and drivers
5. File formats and conventions
6. Games and screensavers
7. Miscellanea
8. System administration commands and daemons


Short answer to how do i see the crontab schedule format is – run

 man 5 crontab 

Per the man page, the time and date fields in order are –

field allowed values
----- --------------
minute 0-59
hour 0-23
day of month 1-31
month 1-12 (or names, see below)
day of week 0-7 (0 or 7 is Sun, or use names)


One of the most commonly used Linux system administration tools is chown, which is part of the coreutils package. It is used to change the user and/or group ownership of a given file or directory. Something to be aware of this tool is, it doesn’t change the ownership of symbolic links, as shown below –

root@linubuvma:/tmp# touch test
root@linubuvma:/tmp# ls -l test
-rw-r--r-- 1 root root 12 Dec 20 08:01 test
root@linubuvma:/tmp# ln -s test sltest
root@linubuvma:/tmp# ls -l sltest
lrwxrwxrwx 1 root root 4 Dec 20 08:01 sltest -> test
root@linubuvma:/tmp# chown daniel:daniel sltest
root@linubuvma:/tmp# ls -l sltest
lrwxrwxrwx 1 root root 4 Dec 20 08:01 sltest -> test

The reason this doesn’t work is in the man page for chown – symbolic links named by arguments are silently left unchanged unless -h is used.” By simply running chown on symbolic link without ‘-h’ option, you are changing the ownership of the target. The ‘-h’ option affects symbolic links instead of any referenced file.

root@linubuvma:/tmp# chown -h daniel:daniel sltest

root@linubuvma:/tmp# ls -l sltest
lrwxrwxrwx 1 daniel daniel 4 Dec 20 08:01 sltest -> test

Though not portable, in some distros

 chown -R 

will recursively change the owernship of all files, including symbolic link files and directories. In my case, ‘chown -R /path/to/file’ works for GNU chown which is part of the ‘GNU coreutils 8.21’ package on Ubuntu 14.04.

How to record your ssh session using screen.

Per the man page – “Screen is a full-screen window manager that multiplexes a physical terminal between several processes (typically interactive shells)”.
Screen is most commonly used to create multiple sessions to remote hosts within a single terminal window or even run multiple commands locally without leaving your shell terminal. For instance, you could be tailing the log file in one session, then run a long process, then ssh into other machine etc. all within a single window.

Screen is the go to tool when setting up a remote connection, such as ssh, and you want to continue your work at any time or from any other host without worrying of a dropped connection.

In this post, I will show you how you can record your bash session.

Installation –

yum install screen        (Debian/Ubuntu)
apt-get install screen    (Redhat/CentOS)

My local environment and the remote host I am sshing to –

daniel@linubuvma:/tmp$ screen -v
Screen version 4.01.00devel (GNU) 2-May-06
daniel@linubuvma:/tmp$ uname -r
3.13.0-106-generic
daniel@linubuvma:/tmp$ cat /etc/issue
Ubuntu 14.04.5 LTS \n \l

daniel@linubuvma:/tmp$ ssh ns2 'uname -r ; cat /etc/issue'
2.6.32-642.6.1.el6.x86_64
CentOS release 6.8 (Final)
Kernel \r on an \m

The ‘-L’ option of screen is used to record your session, the session log is automatically saved in a file named ‘screenlog.n’ in your current directory.

daniel@linubuvma:/tmp$ ls
config-err-hbzs5e  one          ssh-4yheApHRgMBF  ssh-RK7GpeFuzUB8  VMwareDnD    vmware-root-2347660412
gpg-kZux7q         screenlog.0  ssh-BBblvGtb5284  vmware-daniel     vmware-root
daniel@linubuvma:/tmp$ free -m
             total       used       free     shared    buffers     cached
Mem:          3946       2489       1457          6        547       1031
-/+ buffers/cache:        911       3035
Swap:         4092          0       4092
daniel@linubuvma:/tmp$ exit
[screen is terminating]
daniel@linubuvma:/tmp$ 

The whole bash session will be logged in screenlog.0 in this case –

daniel@linubuvma:/tmp$ cat screenlog.0 
daniel@linubuvma:/tmp$ ls
config-err-hbzs5e  one          ssh-4yheApHRgMBF  ssh-RK7GpeFuzUB8  VMwareDnD    vmware-root-2347660412
gpg-kZux7q         screenlog.0  ssh-BBblvGtb5284  vmware-daniel     vmware-root
daniel@linubuvma:/tmp$ free -m
             total       used       free     shared    buffers     cached
Mem:          3946       2489       1457          6        547       1031
-/+ buffers/cache:        911       3035
Swap:         4092          0       4092
daniel@linubuvma:/tmp$ exit
exit
daniel@linubuvma:/tmp$ 

Recording your session of an ssh connection to a remote host is also similar, with ‘-L’ option followed by the command to ssh to remote host.
Option -fn (with no flow-control)
Option -t (title bar name) in this case ‘practice’.

daniel@linubuvma:/tmp$ screen -fn -t practice -L  ssh ns2
Last login: Tue Dec 27 09:46:10 2016 from linubuvma.home.net

[daniel@kauai ~]$ hostname -f
kauai.example.net
[daniel@kauai ~]$ uptime
 10:08:18 up 18 days, 10:02, 14 users,  load average: 0.19, 0.49, 0.64
[daniel@kauai ~]$ exit
[screen is terminating]


daniel@linubuvma:/tmp$ cat screenlog.0 
Last login: Tue Dec 27 09:46:10 2016 from linubuvma.home.net
[daniel@kauai ~]$ hostname -f
kauai.example.net
[daniel@kauai ~]$ uptime
 10:08:18 up 18 days, 10:02, 14 users,  load average: 0.19, 0.49, 0.64
[daniel@kauai ~]$ exit
logout
Connection to ns2 closed.
daniel@linubuvma:/tmp$ 

Additional resources –

https://www.rackaid.com/blog/linux-screen-tutorial-and-how-to/
https://linux.die.net/man/1/screen

Randomly ordering files in a directory with python

I have a playlist file which contains audio files to play. The audio player unfortunately plays the music files in a sequential order, in whatever order they are listed in the playlist file. So occasionally I have to regenerate the playlist file to randomize the audio files order. Here is a simple script that I had to write for this purpose, the core component is the random.shuffle(list) python function –

Create script file as shuffle_files.py –

#!/usr/bin/env python

import os
import random
import sys

music_files=[]

if len(sys.argv) != 2:
  print "Usage:", sys.argv[0], "/path/directory"
else:
  dir_name=sys.argv[1]
  if os.path.isdir(dir_name):
    for file_name in os.listdir(dir_name):
      music_files.append(file_name)
  else:
    print "Directory", dir_name, "does not exist"
    sys.exit(1)
# shuffle list
random.shuffle(music_files)
for item in music_files:
  print os.path.join(dir_name,item)

Run the script by providing a path to a directory with files. Each iteration should list the files in the directory in a different order.
Note – the script does not recurse into the directories, it can be easily modified with os.walk if necessary.

root@svm1010:/home/daniel/scripts# python shuffle_files.py /opt/iotop/iotop
/opt/iotop/iotop/setup.py
/opt/iotop/iotop/README
/opt/iotop/iotop/iotop
/opt/iotop/iotop/iotop.8
/opt/iotop/iotop/NEWS
/opt/iotop/iotop/iotop.py
/opt/iotop/iotop/PKG-INFO
/opt/iotop/iotop/THANKS
/opt/iotop/iotop/sbin
/opt/iotop/iotop/setup.cfg
/opt/iotop/iotop/ChangeLog
/opt/iotop/iotop/.gitignore
/opt/iotop/iotop/COPYING


root@svm1010:/home/daniel/scripts# python shuffle_files.py /opt/iotop/iotop
/opt/iotop/iotop/PKG-INFO
/opt/iotop/iotop/COPYING
/opt/iotop/iotop/iotop
/opt/iotop/iotop/setup.cfg
/opt/iotop/iotop/NEWS
/opt/iotop/iotop/README
/opt/iotop/iotop/.gitignore
/opt/iotop/iotop/setup.py
/opt/iotop/iotop/THANKS
/opt/iotop/iotop/iotop.py
/opt/iotop/iotop/ChangeLog
/opt/iotop/iotop/iotop.8
/opt/iotop/iotop/sbin


root@svm1010:/home/daniel/scripts# python shuffle_files.py /opt/iotop/iotop
/opt/iotop/iotop/THANKS
/opt/iotop/iotop/setup.py
/opt/iotop/iotop/NEWS
/opt/iotop/iotop/README
/opt/iotop/iotop/iotop.8
/opt/iotop/iotop/.gitignore
/opt/iotop/iotop/ChangeLog
/opt/iotop/iotop/sbin
/opt/iotop/iotop/PKG-INFO
/opt/iotop/iotop/iotop
/opt/iotop/iotop/COPYING
/opt/iotop/iotop/iotop.py
/opt/iotop/iotop/setup.cfg

Reference – https://docs.python.org/2/library/random.html?highlight=shuffle#random.shuffle

Ngrep is a very user friendly packet sniffer, basically the “grep” equivalent at the network layer.

Here is a quick way of figuring out the http connections your browser is making even if you are browsing to a secure site, make sure that is the only site you are visiting as the command will capture all port 80 connections.

Installation –

apt-get install ngrep

Let us redirect all traffic ngrep captured to a file –

ngrep -d any -W byline port 80 | tee  /tmp/net_output

Now visit a secure site, say https://cnet.com, you will see nicely formated output

root@lindell:~# ngrep -d any -W byline port 80 | tee  /tmp/output
interface: any
filter: (ip or ip6) and ( port 80 )
####
T 17.31.198.19:33954 -> 72.21.91.29:80 [AP]
POST / HTTP/1.1.
Host: ocsp.digicert.com.
User-Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:50.0) Gecko/20100101 Firefox/50.0.
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8.
Accept-Language: en-US,en;q=0.5.
Accept-Encoding: gzip, deflate.
Content-Length: 83.
Content-Type: application/ocsp-request.
Connection: keep-alive.
..

From here, you can parse the /tmp/output file.

Similarly, you can parse the output file for the type of web server your favorite sites are using. Keep the ngrep command running, and visit all your favorite sites. Note, this works for http only, as https traffic is encrypted, for https only destination IP and port are shown.

In this case, I searched for the ‘Server:’ field in the HTTP response header from the web server. Apparently, nginx seems to be most popular, it is also interesting to see that AmazonS3 storage being used for hosting static content –

root@lindell:~# awk '/Server:/ {print $2}' /tmp/output |sort | uniq -c |sort -nr
    155 nginx.
     40 Apache.
     36 Apache-Coyote/1.1.
     20 Apache/2.2.3
     14 nginx/1.8.1.
      7 AmazonS3.
      6 Akamai
      5 ECS
      5 cloudflare-nginx.
      4 Omniture
      4 ESF.
      3 sffe.
      3 nginx/1.10.2.
      2 Microsoft-IIS/7.5.
      2 gws.
      2 AkamaiGHost.
      1 WildFly/8.
      1 Varnish.
      1 openresty.
      1 NetDNA-cache/2.2.
      1 Cowboy.
      1 ATS.
      1 Apache/2.2.14

References –
http://ngrep.sourceforge.net/usage.html
https://wiki.christophchamp.com/index.php?title=Ngrep

In some cases, you might want to block all users from logging in to the system or just after you login, you want to prevent everyone else from connecting to the server. During server maintenance, this could be helpful or there are use cases where only one actively logged in user has to do some work if the username is a shared account.

Solution – create the /etc/nologin file, and put the text notice as the body of the file. If a user attempts to log in to a system where this file exists, the contents of the nologin file is displayed, and the user login is terminated.

[root@kauai ~]# echo 'System is under maintenance till Dec. 24, 2PM EST.' > /etc/nologin

Now try to login to the server under non super user –

daniel@linubuvma:~$ ssh ns2
System is under maintenance till Dec. 24, 2PM EST.
Connection closed by 192.168.10.103

If your ssh configuration allows it, root user can login to the server though, the root user will still be greeted with the contents of /etc/nologin file though –

daniel@linubuvma:~$ ssh root@ns2
root@ns2's password:
System is under maintenance till Dec. 24, 2PM EST.
Last login: Sat Dec 12 01:11:35 2015 from linubuvma.home.net
[root@kauai ~]# 

Reference – https://docs.oracle.com/cd/E19683-01/806-4078/6jd6cjs3v/index.html