Unix/Linux

• [The Deep, Dark Secrets of Bash]
• [Korn Shell Functions]
• [HANDY AWK ONE-LINERS]

debug a shell

#!/usr/bin/ksh
exec 2>/tmp/mytest

Run a local script on a remote host

ssh user@host 'sh' < your_script.sh

First of all, this command is a redirection: your shell will open the file your_script.sh and feed it as input to the ssh command. ssh, in turn, will tunnel its stdin to the remote command, namely, sh instance. And sh without arguments reads its script from stdin. So we got sh instance, which is launched on remote host, but reads commands from your local file.

Return a list of Oracle databases running on a list of remote servers

#!/bin/ksh

serverlist=`cat /home/tools/ini/system/oracle_servers | sort -n | tr "\n" " "`

if [ -z "${serverlist}" ]
then
    echo "no servers found" && exit 1
fi


for server in ${serverlist}
do
    ssh ${server} "ps -ef | grep [o]ra_pmon_" >/tmp/${server}.dblist
done

for server in ${serverlist}
do
    cat /tmp/${server}.dblist | awk -F_ -v SRV=${server} 'BEGIN {print SRV ":"} {print $NF}' | tr "\n" " "
    echo
done

Check RMAN logfiles for errors from cron every day

00 09 * * * /home/ibmtools/scripts/oracle/dosh 'find /home/ibmtools/logs/rman -name "*online.log" -mtime -1 -exec sed -ne "/^RMAN-/,/^$/p" {} \\; -ls' | mailx
 -s 'RMAN errors last night' [email protected]

Array processing

array=(1 2 3)
unset array[2]
echo ${array[2]}          # null
indices=(${!array[@]})    # create an array of the indices of "array"
size=${#indices[@]}       # the size of "array" is the number of indices into it
size=${#array[@]}         # same
echo ${array[@]: -1}      # you can use slices to get array elements, -1 is the last one, etc.
for element in ${array[@]}; do    # iterate over the array without an index

for index in ${indices[@]}        # iterate over the array WITH an index
do
    echo "Index: ${index}, Element: ${array[index]}"
done

for index in ${!array[@]}         # iterate over the array WITH an index, directly

array+=("new element")    # append a new element without referring to an index
((counter++))             # shorter than ((counter=counter+1)) or ((counter+=1))
if [[ $var == 3 ]]        # you can use the more "natural" comparison operators inside double square brackets
while [[ $var < 11 ]]     # another example
echo ${array[${index}-1]  # math inside an array subscript

Handy aliases

Strip out comments and blank lines from a file

alias strip='grep -Ev '\''^(#|$)'\'''

Does a ps and a grep

alias psg='ps -ef | grep -v $$ | grep -i '

A handy way of listing subdirectories and their files

alias filetree="find . -print | sed -e 's;[^/]*/;|____;g;s;____|; |;g'"

Watch progress of a copy

alias cpProgress="rsync --progress -ravz"

Reboots Linksys router

alias rebootlinksys="curl -u 'admin:password' 'http://192.168.1.2/setup.cgi?todo=reboot'"

Nice one for bash. Colour codes the prompt depending on the outcome of the previous command

bash_prompt_command()
{
    RTN=$?
    prevCmd=$(prevCmd $RTN)
}
PROMPT_COMMAND=bash_prompt_command
prevCmd()
{
    if [ $1 == 0 ] ; then
        echo $GREEN
    else
        echo $RED
    fi
}
if [ $(tput colors) -gt 0 ] ; then
    RED=$(tput setaf 1)
    GREEN=$(tput setaf 2)
    RST=$(tput op)
fi
export PS1="\[\e[36m\]\u.\h.\W\[\e[0m\]\[\$prevCmd\]>\[$RST\]"

Mmmm, to be looked into. Executes remote commands on a unix box using curl.

#/bin/sh
#
# WAG320N-HACK
# Ver. 1.0
# 12/09/2010
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

# Set username and password in the form of "username:password"
# example: "admin:admin"
my_access="admin:admin"

# Parameters test
if [ -z "$1" ]
then
  echo "wag320n-hack.sh: missing remote command"
  echo "usage: wag320n-hack.sh <remote command>"
  echo "example: wag320n-hack.sh /bin/ls -la /usr/sbin"
  echo "Note: always use full path"
  echo ""
  echo "wag320n-hack.sh - Ver. 1.0 - 12/09/2010"
  echo "Licensed under GPL V. 3"
  echo ""
  exit 0
fi

# Get the command
my_command="ping_size="'$('"$@"' 1>&2)'

curl -s -G -u "$my_access" --data-urlencode 'todo=ping_test' --data-urlencode 'this_file=Diagnostics.htm' --data-urlencode 'next_file=Ping.htm' --data-urlencode 'c4_ping_ipaddr=192.168.1.1' --data-urlencode 'ping_timeout=5000' --data-urlencode 'ping_interval=1000' --data-urlencode 'ping_number=1' --data-urlencode "$my_command" http://192.168.1.1/setup.cgi | sed -e '/HTTP\/1.0 200 OK/q'

Use SSH config file

Host server10
  Hostname 1.2.3.4
  IdentityFile ~/.ssh/id_dsa
  user foobar
  Port 30000
  ForwardX11Trusted yes
  TCPKeepAlive yes

then just connect using

ssh server10

A decent sed tutorial

From grymoire.com

A decent korn/bash shell tutorial

From dartmouth.edu Reproduced here just in case it disappears! Advanced shell scripting

trap

     Example Handling Traps With ksh - Discussion of the kill command


EXAMPLE TEMPLATE:


PRODUCT:    HP-UX 11iV1 Version B.11.11
            HP Tru64 V5.1B PK4
            Sun/Solaris SunOS V5.8
            Linux 2.6 kernel


COMPONENT:  ksh


SOURCE:     Philippe Vouters
            Fontainebleau/France


LOW COST HIGH-TECH PRODUCTS:  http://techno-star.fr


OVERVIEW:

The ksh script below shows how to eventually handle traps in the situation 
where someone might try to kill a script by killing individual commands run 
by that script or the entire process group a script is running in. The kill 
command (usually a shell builtin) may be used to send a signal to a process 
group (with the -<pid> syntax) or an individual process. The example ksh 
script below runs /bin/sleep as the foreground process, the example ksh 
scripts immediately returns when the /bin/sleep process has terminated. Most 
signals sent to the shell are ignored until after the foreground process 
terminates. This is in order to avoid creating zombie processes. Therefore a 
kill <pid> on the example ksh script waits for the termination of the 
/bin/sleep process.

The status value $? in the trap refers to the exit status of the command to run 
and therefore is the exit status of the /bin/sleep process. The called function 
in the trap handler shows how to correctly examine the effect of the kill 
command on the shell or it's children.

To examine the value of $? in a trap handler means that you must understand what
it can be set and how different signals delivered to either the shell or the 
foreground process (or the process group) might affect the value of $?.

The example shell script prints $? using echo but it does not perform tests on 
the value of $?. For a complete solution when attempting to trap signals in a 
shell you would also need code that examined the value of $? after the 
foreground process had completed.


*** CAUTION ***

This sample script has been tested using HP-UX B.11.11, HP Tru64 V5.1B PK4, 
SunOS V5.8 and Fedora Core 4 (homed version of Red Hat Linux).  However, we
cannot guarantee its effectiveness because of the possibility of error in 
transmitting or implementing it. It is meant to be used as a template for 
writing your own scripts, and may require modification for use on your system. 


SCRIPT NOTES:

To notice that the ksh script and /bin/sleep share the same process group
identifier (PGID), issue the following commands:

[philippe@victor ~]$ who
philippe :0           Jan 10 10:16
philippe pts/1        Jan 10 21:30 (:0.0)
philippe pts/2        Jan 10 21:30 (:0.0)
[philippe@victor ~]$ tty
/dev/pts/1
[philippe@victor ~]$ ps -j -t pts/2
  PID  PGID   SID TTY          TIME CMD
11072 11072 11072 pts/2    00:00:00 bash
11113 11113 11072 pts/2    00:00:00 ksh
11116 11113 11072 pts/2    00:00:00 sleep

In this case sending kill -INT -11113 will send SIGINT to the process group 
11113. Both of the ksh and sleep processes are contained within this process 
group.

Important Note:

On HP-UX, you have to $ export UNIX95=1 in order to be able to use the
-j option of the ps command.


SCRIPT:

                             COPYRIGHT (C) 2005 BY
                              HEWLETT-PACKARD COMPANY
                                ALL RIGHTS RESERVED.

     THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED AND COPIED
     ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE AND WITH THE INCLUSION
     OF THE ABOVE COPYRIGHT NOTICE.  THIS SOFTWARE OR ANY OTHER COPIES
     THEREOF MAY NOT BE PROVIDED OR OTHERWISE MADE AVAILABLE TO ANY OTHER
     PERSON.  NO TITLE TO AND OWNERSHIP OF THE SOFTWARE IS HEREBY TRANSFERRED.

     THE INFORMATION IN THIS SOFTWARE IS SUBJECT TO CHANGE WITHOUT NOTICE AND
     SHOULD NOT BE CONSTRUED AS A COMMITMENT BY HEWLETT-PACKARD COMPANY.

     HP ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY OF ITS
     SOFTWARE ON EQUIPMENT THAT IS NOT SUPPLIED BY HP.

     NO RESPONSIBILITY IS ASSUMED FOR THE USE OR RELIABILITY OF SOFTWARE
     ON EQUIPMENT THAT IS NOT SUPPLIED BY HEWLETT-PACKARD COMPANY.

     SUPPORT FOR THIS SOFTWARE IS NOT COVERED UNDER ANY HP SOFTWARE
     PRODUCT SUPPORT CONTRACT, BUT MAY BE PROVIDED UNDER THE TERMS OF THE
     CONSULTING AGREEMENT UNDER WHICH THIS SOFTWARE WAS DEVELOPED.

#!/bin/ksh
function handle_signal
{
        print -n "pid $$ recieved $2 "
        if [[ $1 = 0 ]];then
            print but foreground command ended successfully
        else
                if [[ $1 = $3 ]];then
                    print and so did the last foreground command
                else
                    print -n "and the exit status of the last foreground "
                    print command was $1
                fi
        fi
        # Kill our process group and then ourselves with SIGTERM, giving a 
        # pid of 0 sends the signal to our process group. Killing the process
        # group should kill us as well, this assumes that SIGTERM is not 
        # handled by any process in the process group.
        #
        # This code could be replaced with an exit with an exit value that
        # would indicate what the problem was to the caller. That is replace 
        # these two lines with:
        #
        # exit $3
        #
        # or a specific exit code could be used.
        #
        kill -TERM 0
        kill -TERM $$
        }
OS=$(uname -a | awk '{print $1}')
if [[ "$OS" = "Linux" ]]; then
    offset=256
elif [[ ("$OS" = "HP-UX") || 
        ("$OS" = "SunOS") || 
        ("$OS" = "OSF1") ]]; then
   offset=128
fi
trap 'RC=$?; handle_signal $RC SIGINT $offset+2' INT
trap 'RC=$?; handle_signal $RC SIGQUIT $offset+3' QUIT
/bin/sleep 20
echo $?

DNS not working

Ping to an IP address works

ping 74.125.136.103

but this doesn't

ping www.google.com

Check resolv.conf

cat /etc/resolv.conf
nameserver 95.130.132.17
nameserver 95.130.132.18

I had changed internet provider and forgot to update this. Just to set it to the router address and let that do the resolution

nameserver 192.168.1.1

Setup Oracle Enterprise Linux (RedHat) with yum server

You need to download the yum .repo file from the server, as per the steps below. After this, you need to enable a flag in the .repo file as per your operating system version. Having done these two steps, when you run yum install <pkgname> command on your linux box, the Oracle's yum server will be scanned, the dependent & the relevant rpm's will be download and installed for you.

cd /etc/yum.repos.d

To download files here

wget http://public-yum.oracle.com/public-yum-el5.repo

A file named public-yum-el5.repo will be created in your directory Edit this file and enter enabled=1 against the operating systems which is relevant to you

vi public-yum-el5.repo

Next run the yum command

yum install package-name

To change to static IP address

As root:

cd /etc/networks
vi interfaces

replace the line “iface eth0 inet dhcp” with

iface eth0 inet static
address 192.168.1.100
netmask 255.255.255.0
gateway 192.168.1.1

You should also take a look at the file /etc/resolv.conf and check it has a nameserver entry (probably pointing at your default gateway) or direct to your ISP name servers.

nameserver 192.168.1.1

Run a command on lots of servers in parallel

This is a damn fine AIX utility - part of the CSM Distributed Shell.

dsh -a "ls -al /etc/apache2/*conf"

will list the Apache configuration file on all reachable servers (nodes)