AIX System Resource Controller (SRC) Overview

The System Resource Controller (SRC) provides a set of commands and subroutines to make it easier for the system admins/users to create and control subsystems.

What is Subsystem?
A subsystem is any program or process or set of programs or processes that is usually capable of operating independently or with a controlling system.A subsystem is designed as a unit to provide a designated function.

Features of  System Resource Controller (SRC):

• Subsystem is a set of related programs designed to perform one particular function.
• The subsystems can be sub divided into subservers(daemons).
• SRC helps you to manage the whole subsystems and their respective subservers by creating subsystem groups.
• SRC allows us to stop, start, trace, list and refresh subsystems and subservers (daemons).
• SRC is started during the system initialization with a record for /usr/bin/srcmstr daemon in the /etc/inittab file.

Basic Components of SRC:

There are three basic components of SRC

Subgroup –> SubSystem –> SubServer (daemons)

Subserver

A subserver is a program or process that belongs to a subsystem called as deamons as well. 

Eg: sshd,ftpd 

Subsystem

A subsystem can have multiple subservers and is responsible for starting, stopping, and providing status of subservers.

Eg: gated,inetd,named etc.,

Subsystem Group

A subsystem group is a group of any specified subsystems. Grouping subsystems together allows the control of several subsystems at one time. 

Eg: TCP/IP,  Network Information System (NIS), and Network File Systems (NFS).

In the above example 

Subsystem group "tcp/ip" under which there is subsystem "inetd" under which there is a subserver called "ftp" here.

Subsystem Operational  Commands: 

Item	Description
srcmstr	Deamon starts the System Resource Controller
startsrc	Starts a subsystem, subsystem group, or subserver
stopsrc	Stops a subsystem, subsystem group, or subserver
refresh	Refreshes a subsystem
traceson	Turns on tracing of a subsystem, a group of subsystems, or a subserver
tracesoff	Turns off tracing of a subsystem, a group of subsystems, or a subserver
lssrc	Gets status on a subsystem.

Command	Description
lssrc -a	To list the status of all subsystems
lssrc -h node1-a	To list the status of all subsystems  on foreign host node1
lssrc -s inetd	To list the status of the subsystem inetd
lssrc -g tcpip	To get the status of the subsystem group tcpip
startsrc -s inetd	To start the subsystem inetd
startsrc -g tcpip	To start the subsystem group tcpip
stopsrc -s inetd	To stop the subsystem inetd (If process is under srcmstr. ie PPID of process=PID of srcmstr)
stopsrc -g tcpip	To stop the subsystem group tcpip
refresh -s nfsd	To refresh nfsd subsystem
refresh -g tcpip	To refresh tcpip subsystem group
lssrc -p [PID of process]	To get  status of the subsystem by process ID
kill  [PID of process]	To kill a process that not started by srcmstr

Subsystem Config Commands:

mkssy ==> Create Subsystem
chssys  ==> Chnage or modify Subsystem Parameters
rmssys  ==> Remove Subsystem

# mkssys -p /usr/sbin/sshd \   /* Absolute path to the subsystem executable
                                  program. */
         -s sshd_adm \         /* Name that uniquely identifies the subsys. */
         -u 0 \                /* User id for the subsystem. */
         -a "-D -f /etc/ssh/sshd_config_adm" \   /* Arguments that must be
                                                    passed to the command. */
         -e /dev/console \     /* Where the subsystem standard error data is
                                  placed. */
         -i /dev/console \     /* Where the subsys. standard input is routed. */
         -o /dev/console \     /* Where the subsys. standard output is placed. */
         -R \                  /* Subsystem is restarted if the subsystem stops
                                  abnormally. */
         -Q \                  /* Multiple instances of the subsystem are not
                                  allowed to run at the same time. */
         -S \                  /* Subsystem uses the signals communication
                                  method. */
         -f 9 \                /* Signal sent to the subsystem when a forced
                                  stop of the subsystem is requested. */
         -n 15 \               /* Signal sent to the subsystem when a normal
                                  stop of the subsystem is requested. */
         -E 20 \               /* Execution priority of the subsystem. */
         -G ssh \              /* Subsystem belongs to the group specified. */
         -d \                  /* Inactive subsystems are displayed when the
                                  lssrc -a command request is made. */
         -w 20                 /* Time, in seconds, allowed to elapse between a
                                  stop cancel (SIGTERM) signal and a subsequent
                                  SIGKILL signal. */
Check the service's configuration:

# lssrc -S -s sshd_adm                
#subsysname:synonym:cmdargs:path:uid:auditid:standin:standout:standerr:action:multi:contact:svrkey:svrmtype:\
 priority:signorm:sigforce:display:waittime:grpname:
sshd_adm::-D -f /etc/ssh/sshd_config_adm:/usr/sbin/sshd:0:0:/dev/console:/dev/console:/dev/console:-R:-Q:-S:0:0:\
20:15:9:-d:20:ssh:

# odmget -q subsysname=sshd_adm SRCsubsys

SRCsubsys:
        subsysname = "sshd_adm"
        synonym = ""
        cmdargs = "-D -f /etc/ssh/sshd_config_adm"
        path = "/usr/sbin/sshd"
        uid = 0
        auditid = 0
        standin = "/dev/console"
        standout = "/dev/console"
        standerr = "/dev/console"
        action = 1
        multi = 0
        contact = 2
        svrkey = 0
        svrmtype = 0
        priority = 20
        signorm = 15
        sigforce = 9
        display = 1
        waittime = 20
        grpname = "ssh"

Read More

Default ports of ITM components

1) PORT 1920:

By default, a http port 1920 is allocated during starting of the first ITM component. This port is used to serve service console request as well as the TEPS and SOAP server request. if this port is not available, a random port is allocated and used for http request.

The owner of the base HTTP port 1920 will redirect the calls to the random http port allocated by the other ITM components. The same can happen to the HTTPS port 3661.

Note that the allocated random http port is bound to 1920, so the http request is served using a random port as well as 1920

The random http and https ports allocated by the ITM components can be identified from the RAS1 log of the ITM component

(4FEDD526.0015-32B4:kdhslqm.c,349,"add_listener") listening: ip.tcp.http:59029
(4FEDD526.0027-32B4:kdhslqm.c,349,"add_listener") listening: ip.ssl.https:59031
(4FEDD526.0013-32B4:kdhslqm.c,349,"add_listener") listening: ip6.tcp.http:59028
(4FEDD526.0021-32B4:kdhslqm.c,349,"add_listener") listening: ip6.ssl.https:59030

2) PORT 3661:

by default, a https port 3661 is allocated for secure http request during starting of the first ITM component. This port is used to serve service console request as well as the TEPS and SOAP server request. if this port is not available, a random port is allocated and used for https request.

3) PORT 4096:

A basic services port is allocated which is based on ITM port allocation algorithm
(well-known port + count * 4096)

where the well-known port is the port number assigned to the monitoring server . The default port number assigned to monitoring server is 1918.

Let's say, if the well-known port assigned to monitoring server is 1918, the first started monitoring agent gets 1918+4096=6014, if the port 6014 is busy then 1918 + 2 * 4096 = 10110 port will be allocated to the monitoring agent and so on.. if the port 6014 is RESERVED, then the agent won't start and fail

4)PORT 15001:

In addition TEPS uses 15001 for its own purpose.

NOTE : all the above ports can be controlled using POOL parameter and the explanation of POOL parameter is not the scope of this technote

5) PORT 9999:

By default, eclipse help server uses port 9999. This can be changed by reconfiguring the eclipse help server

6) Loopback Ports:

Also, whenever the ITM components including agent started, there are few outbound loopback ports allocated. Below is the some example for loopback ports:

127.0.0.1.1052
127.0.0.1.1053
127.0.0.1.1054

7) PORT 1978:

Default port 1978 is used by the remote deployment process

NOTE: The loopback and the default remote deployment ports cannot be directed to specific port range

Read More

Where does my space gone in AIX/Linux filesystem ?

One of my friend got a situation where in she is seeing 9 GB allocated to one of the filesystems which is 100%utilized  but actual usage is 4GB  when verified with "du" command

#df -k  /mytest
Filesystem 1024-blocks Free %Used Iused %Iused Mounted on
/dev/mytestlv  9216 9216 100% 48804 12% /mytest 

She was wondering where does the other 5GB gone . 

Reason: 

This situation happens  when a  process is opening a file and dumping data into it and the file is removed while said process still has file open.So called process still holds that file space even file deleted.

How to Rectify ?

At first you need to check  what are all the  processes using  a particular filesystem using "fuser" command.

# fuser -c /mytest
/mytest:  2567 4006c 6548c 8657

You need to kill the above process if you want to free up the space.

Note: You need to inform the respective application owner/support team and take the application down time if this file-system is used by any-application.

How to Kill the proceses ?

# fuser -kc /mytest 

This  will kill  all the processes and  space will be freed up.

Check the space now 

#df -k  /mytest
Filesystem 1024-blocks Free %Used Iused %Iused Mounted on
/dev/mytestlv  9216 4096 45% 48804 12% /mytest 

Read More

MS Word Secret Code

I think most of you all know about this but I wanna  remind again.

Do you want a Word document full of text without typing it all?

This is useful specially for print test purpose.

open Microsoft Word in RUN: type winword and press enter.

open a new Word document and type in

=rand(3,9)

then press''enter'

3 ==> number of paragraphs
9 ==>  number of sentences for paragraph.

Its your choice  to choose the number of paragraphs and sentences.

After Enter  you would see this 

Read More

Not enough free space to shrink the file system issue in AIX

Recently got an issue in reducing jfs2 filesystem  with osverion 6.1 and have enough space to reduce filesystem.

root@umaix /tmp>df -g /orafs1
Filesystem    GB blocks      Free %Used    Iused %Iused Mounted on
/dev/oralv1   100.00    75.00   25%      555     1%  /orafs1

root@umaix /tmp>chfs -a size=-15G /orafs1
chfs: There is not enough free space to shrink the file system.

This issue will occur whenever you try to reduce big chunk of data ( in this case 15GB) that may not be contiguous in the file-system because you have files scatted everywhere.

Try   the following methods one by one until your issue fixed

1. Try to defrag the FS:

#defragfs -s /orafs1

2. Reduce in smaller chunks:

If you still can't reduce it after this. Try reducing the filesystem  in smaller chunks. Instead of 15G at a time, try reducing 1 or 2 gigs. Then, repeat the operation.

3. Check the processes:

Sometimes processes open big files and use lots of temporary space in those filesystem.
You could check processes/applications running against the filesystem and stop them temporarily, if you can.

#fuser -cu[x] <filsystem>

4. Move the large files and try shrink

Try looking for files large using the find cmd and move them out temporarily, just to see if we can shrink the fs without them:

#find /<filesystem> -xdev -size +2048 -ls|sort -r +10|pg

Finally the last method, the alternative approach if any one of above methods are not working then go for filesystem recreation.

==> You should be very care full , need to take fs backup and as well as approach application before removing the filelsystem.

5) Recreate filesystem:

• - Take databackup of the fielsystem  ( very Important,dont skip this )
  Either you can take using your backup tools like TSM / netbackup or move data to a temporary   directory


• - Remove the  filesystem  (  #rmfs /orafs)
• - Create the filesystem again
   #mklv -y oralv1 -t jfs2 oravg 600  ( in this case we need 75GB and pp size is 128)
   #crfs -v jfs2 -d oralv1 -m /orafs1 -A yes  (create orafs1 filesystem)


• - Restore data to the filesystem
• - Verify fs size

root@umaix /tmp>df -g /orafs1
Filesystem    GB blocks      Free %Used    Iused %Iused Mounted on
/dev/oralv1   75.00    50.00   33%      555     1%  /orafs1

Read More

How to mirror VIOS Boot Disk?

Here is the procedure to mirror VIOs boot disk.

# lspv
NAME             PVID                 VG               STATUS
hdisk0           00c122d4341c6e62     rootvg           active
hdisk1           00cd55a4fg6b676f     None
hdisk2           00c5524409a99b77     None

Here hdisk0 is rootvg disk , now we need to check free disk.
you can use lspv -free command to check the un-mapped free disks.

$ lspv -free
NAME            PVID                                SIZE(megabytes)
hdisk1         00cd55a4fg6b676f                     256000
hdisk2         00c5524409a99b77                     256000

So In this case, hdisk1 is free and un-mapped . So we're going to use hdisk1 to mirror with hdisk0.

Add hdisk1 into rootvg:

# extendvg rootvg hdisk1 0516-1254 extendvg: Changing the PVID in the ODM.

Now mirror the disk but defer the automatic reboot:

$ mirrorios -defer hdisk1

Now check the boot list:

$ bootlist -mode normal -ls
hdisk0 blv=hd5 pathid=0

We only have hdisk0 at the moment.  Need to add hdisk1 into this:

$ bootlist -mode normal hdisk0 hdisk1

Check that worked:

$ bootlist -mode normal -ls
hdisk0 blv=hd5 pathid=0
hdisk1 blv=hd5 pathid=0

You now have a mirrored rootvg. 

Read More

Tar Files extraction Unix / Linux

Q: How can I extract  specific file from a tar ball ?

Tar or Tar ball is a single file  bundled with files &/ directories. First  we will discuss about general extraction of a files from tar ball.

Unpack or extract a tar file :

To unpack or extract a tar file, type:

tar -xvf myfile.tar

some times to save more space and bandwidth , we  need compress the tar balls using compression techniques like gzip or bzip2.

To unzip and extract  those tar files, type as below 

For  .tar.gz files 

tar -xzvf myfile.tar.gz

For .tar.bz2

tar -xjvf myfile.tar.bz2

Where,
-x : Extract a tar ball.
-v : Verbose output or show progress while extracting files.
-f : Specify an archive or a tarball filename.
-j : Decompress and extract the contents of the compressed archive created by bzip2 program (tar.bz2 extension).
-z : Decompress and extract the contents of the compressed archive created by gzip program (tar.gz extension).
Now here comes our main purpose ,  to extract a specific file  from a tar file.

Extract Specific file from a tar ball:

To extract a single file called myfile1.txt, enter:

tar -xvf file.tar myfile1.txt
tar -xzvf file.tar.gz myfile1.txt
tar -xjvf file.tar.bz2 myfile1.txt

You can also specify path such as home/um/myfile2.txt, enter:

tar -xvf file.tar home/um/myfile2.txt
tar -xzvf file.tar.gz home/um/myfile2.txt
tar -xjvf file.tar.bz2 home/um/myfile2.txt

How to Extract a Single Directory?

To extract a single directory called /home/um, enter:

tar -xvf file.tar home/um
tar -xzvf file.tar.gz home/um
tar -xjvf file.tar.bz2 home/um

Sample O/P:
home/um/
home/um/ddl/
home/um/ddl/default
home/um/ddl/bin/config.conf
home/um/ddl/daemon.conf
home/um/ddl/config/system.sh

Read More

MobaXterm free Xserver and SSH client

Recently when I am browsing internet  for Unix connectivity tools I came across a wonderful tool called MobaXterm. As per the MobaXterm team  its "an enhanced terminal for Windows with an X11 server, a tabbed SSH client and several other network tools for remote computing (VNC, RDP, telnet, rlogin). MobaXterm brings all the essential UNIX commands to Windows desktop, in a single portable exe file which works out of the box".

Downaload:There are two different versions available, one is free version and other is Pro.
I think free version is good enough to accommodate our daily tasks. If you still  feel you need  more then  go for Pro.

Download Link

The following are  the key features mentioned by them.

Key Features:

• Free X server fully configured (based on Xorg)
• Tabbed terminal based on PuTTY / MinTTY with antialiased fonts and macro support
• Easy DISPLAY exportation from any remote host
• X11-Forwarding capability in OpenSSH
• Several Unix/Linux commands based on Cygwin (rsync, wget, sed, awk, grep, cd, ls, cat, cp, ...)
• A session manager with all the network clients you need: RDP, VNC, SSH, telnet, rsh, FTP, SFTP and XDMCP
• Program without installation that you can start from an USB stick
• Light and portable application, packaged in a single executable

I am really enthralled by the features its offering.Here are few snippets.

1) Inbuilt Unix Command Support :
When you open MobaXterm  it will automatically  open session which supports many of Unix/Linux based commands  (ls, cd, grep, awk, tail, cut, sed, wget, rsync, telnet, ssh, rlogin, rsh etc ) based on Cygwin. You can further add packages using this link plugins.

2)Tabbed Sessions:
When ever you open a session , it will be open as tabbed session, just like Putty Connection Manager.
You can also save your session along with usernames and passwords

3)X11-Forwarding
The SSH connection configured with automatic X-11 forwarding.You can use this as fully configured "X server " and also supports OpenGL.

4) Automatic SFTP Support:
A SFTP session is automatically opened when you open a ssh session on leftside. You can  download or upload files directly with a drag and drop from/to the remote server.You  can also  edit And save the files with one click open it in MobiXeditor or  which ever the text editor on your system.

5) Remote Windows Connections using RDP: 
  You can access the remote windows machines using RDP adding as connection.

6) Split-mode Terminal:
you can display up to 4 terminals at the same time in the main window. It is very useful for monitoring 4 remote computers or for comparing outputs of 2 terminals.

7) Multiple Sessions Support:
It will support various session types like SSH,RSH,Telnet,Xdmcp,RDP,VNC, SFTP,FTP many more.

8) Muti Execution Mode:
There us a facility just klike PuttyCM, You can run one single command in all the open sessions.But be more cautious when you are using this facility, need to be careful on which sessions you are running this.

---
There are lot more features and further information you can refer the following link.
more features link

Read More

Manually Installing PHP in Linux

Before installing php,  we need to install apache.The most recent version of Apache HTTP Server may be obtained from >> Apache Download.

1) Download  & Unpack Apache HTTP server Package:

Download and unpack Apache http server package  from the location listed above, and unpack it.
Download Link: Apache Download

gzip -d httpd-2_x_NN.tar.gz
tar -xf httpd-2_x_NN.tar

2) Download  & Unpack PHP source Package:

Download Link: PHP Dowanlod

gunzip php-NN.tar.gz
tar -xf php-NN.tar

3) Build and install Apache:

cd httpd-2_x_NN
./configure --enable-so
make
make install

4) Start  & Start Apache:

/usr/local/apache2/bin/apachectl start

stop the  apache to configure php

/usr/local/apache2/bin/apachectl stop

5) Configure & Build  PHP Package:

cd ../php-NN
./configure --with-apxs2=/usr/local/apache2/bin/apxs --with-mysql
make
make install

6)Setup your php.ini

cp php.ini-development /usr/local/lib/php.ini

You may edit your .ini file to set PHP options. If you prefer having php.ini in another location, use --with-config-file-path=/some/path in step 5.

If you instead choose php.ini-production, be certain to read the list of changes within, as they affect how PHP behaves.

7) Edit your httpd.conf to load the PHP module:

LoadModule php5_module modules/libphp5.so

8) Tell Apache to  parse PHP extensions:

If you instead choose php.ini-production, be certain to read the list of changes within, as they affect how PHP behaves.
let's have Apache parse .php files as PHP. Add to httpd.conf file.

<FilesMatch \.php$>
    SetHandler application/x-httpd-php
</FilesMatch>

Or, if we wanted to allow .php, .php2, .php3, .php4, .php5, .php6, and .phtml files to be executed as PHP, but nothing else, we'd use this:

FilesMatch "\.ph(p[2-6]?|tml)$">
    SetHandler application/x-httpd-php
</FilesMatch>

And to allow .phps files to be handled by the php source filter, and displayed as syntax-highlighted source code, use this:

<FilesMatch "\.phps$">
    SetHandler application/x-httpd-php-source
</FilesMatch>

9) Start Apache:

/usr/local/apache2/bin/apachectl start

OR

service httpd restart

That’s all.

Read More

WEBMIN- Managing Unix Systems Graphically

What is Webmin?

Webmin is a web-based interface for system administration for Unix. Using any modern web browser, you can setup user accounts, Apache, DNS, file sharing and much more.

Demo:

http://webmin-demo.virtualmin.com/   login: demo &  password: demo.

Download Link:

http://www.webmin.com/download.html 

How to Install:

Install on RedHat/CentOS/Fedora:

If you are using the RPM version of Webmin, first download the file from the downloads page , or run the command :

[root@UMLinux1 ~]# wget http://prdownloads.sourceforge.net/webadmin/webmin-1.710-1.noarch.rpm

and then run the command

[root@UMLinux1 ~]# rpm -U webmin-1.710-1.noarch.rpm

The rest of the install will be done automatically to the directory /usr/libexec/webmin, the administration username set to root and the password to your current root password. You should now be able to login to Webmin at the URL http://localhost:10000/.Or if accessing it remotely, replace localhost with your system's IP address.

If you want to connect from a remote server and your system has a firewall installed, see this page for instructions on how to open up port 10000.

Install on Debian:

If you are using the DEB version of webmin, first download the file from the downloads page , or run the command :

[root@UMLinux1 ~]# wget http://prdownloads.sourceforge.net/webadmin/webmin_1.710_all.deb

then run the command :

[root@UMLinux1 ~]# dpkg --install webmin_1.710_all.deb

The install will be done automatically to /usr/share/webmin, the administration username set to root and the password to your current root password. You should now be able to login to Webmin at the URL http://localhost:10000/. Or if accessing it remotely, replace localhost with your system's IP address.

How to Stop& Start Webmin Services:

In order to start the Webmin service on CentOS (Linux) you will need to issue the following command:

[root@UMLinux1 ~]# service webmin start

You can check to make sure that Webmin is running by issuing the following command:

[root@UMLinux1 ~]# service webmin status
Webmin (pid 1729) is running
[root@UMLinux1 ~]#

If you wish to configure your server to ensure that the Webmin service is started at boot time you can issue the following command:

[root@UMLinux1 ~]# chkconfig --level 3 webmin on

To verify that Webmin will start at boot, issue the following command:

[root@UMLinux1 ~]# chkconfig --list webmin
webmin 0:off 1:off 2:off 3:on 4:off 5:off 6:off
[root@UMLinux1 ~]#

In the previous listing, Webmin is listed to start in run level 3, which is the default run level that the dedicated servers boot into.

Read More

Run VIO commands from the HMC using "viosvrcmd" without VIOs Passwords

Recently  we got a situation where  in we don't know the passwords of  either padmin/root of VIOS  but need to run commands in VIOs.

Found an interesting command  in HMC  called "viosvrcmd",which will enble us to run commands on VIOs through HMC.

viosvrcmd -m managed-system {-p partition-name | --id partition-ID} -c "command" [--help]

Description: viosvrcmd issues an I/O server command line interface (ioscli) command to a virtual I/O server partition.

The ioscli commands are passed from the Hardware Management Console (HMC) to the virtual I/O server partition over an RMC session.

RMC does not allow interactive execution of ioscli commands.

-m    VIOs managed system name

-p    VIOs hostname

--id  The partion ID of the VIOs

Note:You must either use this option to specify the ID of the partition, or use the  -p option to specify the partition's name. The --id and the -p options are mutually exclusive.

-c    The I/O server command line interface (ioscli) command to issue to the virtual I/O      server partition.

Note: Command must be enclosed in double quotes. Also, command cannot contain the      semicolon (;), greater than (>), or vertical bar (|) characters.

--help  Display the help text for this command and exit.

Here is an example:

hscroot@umhmc:~> viosvrcmd -m umfrm570 -p umvio1 -c "ioslevel"
2.2.0.0

Since  we can't give the ; or > or |  in the command , if you need to process the output using filters , you can use that after "".

hscroot@umhmc:~> viosvrcmd -m umfrm570 -p umvio1 -c "lsdev -virtual" | grep vfchost0
vfchost0         Available   Virtual FC Server Adapter

What if  you want to run  command as root (oem_setup_env) ,  

got a method from internet

hscroot@umhmc:~> viosvrcmd -m umfrm570 -p umvio1 -c "oem_setup_env
> whoami"
root

You can  run in one shot like below

hscroot@umhmc:~> viosvrcmd -m umfrm570 -p umvio1 -c "oem_setup_env\n whoami"
root

If you need to run multiple commands , you can use them by assiging the commands to a variable and call the variable in place of the command parameter.

hscroot@umhmc:~>command=`printf  "oem_setup_env\nchsec -f /etc/security/lastlog -a unsuccessful_login_count=0 -s padmin"`

hscroot@umhmc:~>viosvrcmd -m umfrm570 -p umvio1 -c "$command"

Read More

Expressions used with if condition in shell scripts

1-1. Primary Expressions

Primary	Meaning
[ -a FILE ]	True if FILE exists.
[ -b FILE ]	True if FILE exists and is a block-special file.
[ -c FILE ]	True if FILE exists and is a character-special file.
[ -d FILE ]	True if FILE exists and is a directory.
[ -e FILE ]	True if FILE exists.
[ -f FILE ]	True if FILE exists and is a regular file.
[ -g FILE ]	True if FILE exists and its SGID bit is set.
[ -h FILE ]	True if FILE exists and is a symbolic link.
[ -k FILE ]	True if FILE exists and its sticky bit is set.
[ -p FILE ]	True if FILE exists and is a named pipe (FIFO).
[ -r FILE ]	True if FILE exists and is readable.
[ -s FILE ]	True if FILE exists and has a size greater than zero.
[ -t FD ]	True if file descriptor FD is open and refers to a terminal.
[ -u FILE ]	True if FILE exists and its SUID (set user ID) bit is set.
[ -w FILE ]	True if FILE exists and is writable.
[ -x FILE ]	True if FILE exists and is executable.
[ -O FILE ]	True if FILE exists and is owned by the effective user ID.
[ -G FILE ]	True if FILE exists and is owned by the effective group ID.
[ -L FILE ]	True if FILE exists and is a symbolic link.
[ -N FILE ]	True if FILE exists and has been modified since it was last read.
[ -S FILE ]	True if FILE exists and is a socket.
[ FILE1 -nt FILE2 ]	True if FILE1 has been changed more recently than FILE2, or if FILE1 exists and FILE2 does not.
[ FILE1 -ot FILE2 ]	True if FILE1 is older than FILE2, or is FILE2 exists and FILE1 does not.
[ FILE1 -ef FILE2 ]	True if FILE1 and FILE2 refer to the same device and inode numbers.
[ -o OPTIONNAME ]	True if shell option "OPTIONNAME" is enabled.
[ -z STRING ]	True of the length if "STRING" is zero.
[ -n STRING ] or [ STRING ]	True if the length of "STRING" is non-zero.
[ STRING1 == STRING2 ]	True if the strings are equal. "=" may be used instead of "==" for strict POSIX compliance.
[ STRING1 != STRING2 ]	True if the strings are not equal.
[ STRING1 < STRING2 ]	True if "STRING1" sorts before "STRING2" lexicographically in the current locale.
[ STRING1 > STRING2 ]	True if "STRING1" sorts after "STRING2" lexicographically in the current locale.
[ ARG1 OP ARG2 ]	"OP" is one of -eq, -ne, -lt, -le, -gt or -ge. These arithmetic binary operators return true if "ARG1" is equal to, not equal to, less than, less than or equal to, greater than, or greater than or equal to "ARG2", respectively. "ARG1" and "ARG2" are integers.

Expressions may be combined using the following operators, listed in decreasing order of precedence:

1-2. Combining expressions

Operation	Effect
[ ! EXPR ]	True if EXPR is false.
[ ( EXPR ) ]	Returns the value of EXPR. This may be used to override the normal precedence of operators.
[ EXPR1 -a EXPR2 ]	True if both EXPR1 and EXPR2 are true.
[ EXPR1 -o EXPR2 ]	True if either EXPR1 or EXPR2 is true.

The [ (or test) built-in evaluates conditional expressions using a set of rules based on the number of arguments. Just like the if is closed with fi, the opening square bracket should be closed after the conditions have been listed.

Read More

How to change system hostname in Linux ?

Recently  we got a request from one of our  visitor to post article related to hostname change in Linux operating systems. I am going to cover this  now.

There are two general way to do this 

1)  Temporary 

2)  Permanent

First we go and learn about how to check host name (system name) of the server.

Use "hostname" command to list the system system name.

[root@umser1 ~]# hostname umser1.unixmantra.com [root@umser1 ~]#

    -s, --short              short host name

    -a, --alias               alias names

    -i, --ip-address      addresses for the hostname

    -I, --all-ip-addresses all addresses for the host

    -f, --fqdn, --long    long host name (FQDN)

    -A, --all-fqdns        all long host names (FQDNs)

    -d, --domain           DNS domain name

    -y, --yp, --nis          NIS/YP domainname

    -F, --file                  read hostname or NIS domainname from given fil

In Cent OS  we have another command additionally 

[root@umser1 ~]# sysctl kernel.hostname

kernel.hostname = umser1.unixmantra.com

[root@umser1 ~]#

Change the hostname on a running system (Temporarily) :

This is pretty simple  

#hostname  new-name

will set the hostname of the system to  new-name. This is active right away and will remain like that until the system will be rebooted (because at system boot it will set this from some particular file configurations – see bellow how to set this permanently). You will most probably need to exit the current shell in order to see the change in your shell prompt.

How Do I Change Hostname Permanently?

For Debian  Systems:

Debian based systems use the file /etc/hostname to read the hostname of the system at boot time and set it up using the init script /etc/init.d/hostname.sh

# /etc/hostname

umser2.unixmantra.com

So on a Debian based system we can edit the file /etc/hostname and change the name of the system and then run:

/etc/init.d/hostname.sh start

to make the change active. The hostname saved in this file (/etc/hostname) will be preserved on system reboot (and will be set using the same script we used hostname.sh).

For Redhat/Fedora/Cent OS Systems:

As you know if you need  changes to be  permanent   you need to hard-code the relevant configuration files.

To make the hostname name permanent in  RH variants ,you must edit the /etc/sysconfig/network file to change  "HOSTNAME" value to your new hostname.

#vi /etc/sysconfig/network

NETWORKING=yes

HOSTNAME="umser2.unixmantra.com"

GATEWAY="192.168.1.1"

GATEWAYDEV="eth0"

FORWARD_IPV4="yes"

Verification:

Open new session and  there you go ,we can  see our  new hostname

[root@umser2 ~]# hostname

umser2.unixmantra.com

[root@umser2 ~]#

Read More

Getting "Server refused to allocate pty" upon login attempt

Problem(Abstract)

You are unable to log into AIX because the maximum number of pseudo-terminals have already been allocated.

Symptom

An attempt to log into AIX via telnet or ssh results in this error:

"Server refused to allocate pty"

- You have increased the maximum number of ptys but you still see the problem.
- Each time you log in, the pty number increases and the pty numbers are not getting released and re-used.

Diagnosing the problem

The symptoms may indicate that there is an application that is holding on to ptys and not releasing it.

Try using the 'fuser' command to find the culprit application, like this:

# cd /dev/pts
# fuser *

The 'fuser' command will list all PIDs associated with each pty device.

If there is a process that is not releasing its ptys, you will see its PID occur many times in the fuser output above

Resolving the problem

Restarting the application that you diagnosed above should release all the ptys held by that application. Contact the application vendor support to see if there is a patch or configuration for the problem.

Read More

PowerHA/HACMP Moving Resource Group (RG) one node to other

We are going to discuss about the resource group (RG) movement one node to other in PowerHA.
Here are the steps

1) Extending PATH vairable with cluster paths

Sometimes cluster paths are not included in default path ,run below command incase if you are not able to run commands directly.

export PATH=$PATH:/usr/es/sbin/cluster:/usr/es/sbin/cluster/utilities:/usr/es/sbin/cluster/sbin:/usr/es/sbin/cluster/cspoc

2) Check the cluster services are up  or not in destination node

#clshowsrv -v
Status of the RSCT subsystems used by HACMP:
Subsystem         Group            PID          Status
 topsvcs          topsvcs          278684       active
 grpsvcs          grpsvcs          332026       active
 grpglsm          grpsvcs                       inoperative
 emsvcs           emsvcs           446712       active
 emaixos          emsvcs           294942       active
 ctrmc            rsct             131212       active

Status of the HACMP subsystems:
Subsystem         Group            PID          Status
 clcomdES         clcomdES         204984       active
 clstrmgrES       cluster          86080        active

Status of the optional HACMP subsystems:
Subsystem         Group            PID          Status
 clinfoES         cluster          360702       active

3) Check the availability of resource group

# clRGinfo
-----------------------------------------------------------------------------
Group Name     Type           State      Location
-----------------------------------------------------------------------------
UMRG1            non-concurrent OFFLINE    umhaserv1
                                ONLINE     umhaserv2
#

4) Move the resourcegroup by using below command

==>  clRGmove -g <RG> -n  <node> -m

# clRGmove -g UMRG1 -n umhaserv1 -m
Attempting to move group UMRG1 to node umhaserv1.
Waiting for cluster to process the resource group movement request....
Waiting for the cluster to stabilize..................
Resource group movement successful.
Resource group UMRG1 is online on node umhaserv1.

You can use smitty path also

smitty cl_admin => HACMP Resource Group and Application Management => Move a Resource Group to Another Node / Site

5) Verify the RG movement

# clRGinfo
-----------------------------------------------------------------------------
Group Name     Type           State      Location
-----------------------------------------------------------------------------
UMRG1          non-concurrent   ONLINE     umhaserv1
                                OFFLINE    umhaserv2
#

Read More

Editing the /etc/inittab File in Maintenance Mode

Problem(Abstract)

This technote describes a technique for creating a minimal /etc/inittab file if no other tools are available.

Symptom

System hangs or crashes at boot time.

Cause

A bad entry in the /etc/inittab is keeping the system from booting properly.

Resolving the problem

Ordinarily if there is a problem with one or more entries in the /etc/inittab the preferred method of editing it is:

1. Boot into Maintenance Mode off AIX install CDs, mksysb, or NIM

2. Access the rootvg and start a shell with the filesystems mounted.

3. Edit /etc/inittab down to a minimum 3 lines:

init:2:initdefault:
brc::sysinit:/sbin/rc.boot 3 >/dev/console 2>&1 # Phase 3 of system boot
cons:0123456789:respawn:/usr/sbin/getty /dev/console

In cases where the rootvg filesystems cannot be mounted automatically (for example the CD media is a different Technology Level than what exists on hard disk; or the filesystems for some reason won't automatically mount), commands such as the "vi" editor won't be available to edit the inittab.

In this case a hard-luck method can be used to create a minimal inittab.

1. Boot into Maintenance Mode and choose Option 2 "Access rootvg and start a shell before mounting filesystems".

2. Once in Maintenance Mode, fsck all rootvg filesystems necessary:

# fsck /dev/hd1
# fsck /dev/hd2
# fsck /dev/hd3
# fsck /dev/hd4
# fsck /dev/hd9var

3. Mount root on a temporary mount point:

# mount /dev/hd4 /mnt

4. Copy the bad inittab to a backup:

# cd /mnt/etc
# mv inittab inittab.bad

5. Use grep to create a minimal new inittab:

# grep "init:" inittab.bad > inittab (adds both the init: and brc: entries)
# grep "^cons:" inittab.bad >> inittab (adds the cons: entry)

6. Reboot using the new inittab:
# sync; sync; sync
# cd /
# umount /mnt

power cycle the system from the front panel or HMC

Read More

How to enable the Name Service cache Daemon (NSCD)

Question

How do you enable NSCD to improve the performance of the hostname, password, name and group lookup which is frequently being done by IBM Rational ClearCase?

Cause

By enabling the Name Service cache Daemon (NSCD) of the operating system, a significant performance improvement can be achieved when using naming services like DNS, NIS, NIS+, LDAP.

Answer

Benefit of name service cache daemon (NSCD) for ClearCase

Example:

WithoutNSCD:

[user@host]$ time cleartool co -nc "/var/tmp/file"
Checked out "/var/tmp/file" from version "/main/10".
real    0m3.355s
user    0m0.020s
sys     0m0.018s

With NSCD

[user@host]$ time cleartool co -nc "/var/tmp/file"
Checked out "/var/tmp/file" from version "/main/11".
real    0m0.556s
user    0m0.021s
sys     0m0.016s

Enabling NSCD

Solaris:
/etc/init.d/nscd start

Linux
service nscd start

AIX:
startsrc -s netcd

Note: In addition to having nscd started it is mandatory to be sure this service will be started after a reboot. For instance on Red Hat and SuSE you can run:

chkconfig nscd  on

For more details on how to configure and or enable NSCD refer to your respective operating system vendor's manpage.

Note that this service is not yet available on HP-UX platforms.

Read More

Howto fix delay in SSH Login

Have you ever faced  login delays  when you tried to connect to the Linux systems, if yes this is happening due to  reverse DNS look-up  query that is been made to DNS Server.

We can fix this issue as mentioned below steps:

1) Take /etc/ssh/sshd_config  backup

# cp -p /etc/ssh/sshd_config /etc/ssh/sshd_config.`date '+%m-%d-%Y_%H:%M:%S'`

2) Edit  /etc/ssh/sshd_config  on sshd  Server

vi /etc/ssh/sshd_config

  And add this DNS option to the file:

  UseDNS no

3) Now add the following line to your /etc/resolv.conf
   options single-request-reopen 4) Restart ssh daemon
# service sshd restart
Sometimes adding the client's net address to the server's /etc/hosts can fix this issue  which is an alternative method. 

Read More

Install SNMP Service on RHEL or CentOS

Install SNMP Service on RHEL or CentOS

In this article  we are going to learn  how to install and start  SNMP service in RHEL/CentOS.

We need  to have  net-snmp rpm package installed on the servers , generally it would come with repository.

1. Install net-snmp with yum:

#yum install net-snmp

[root@umserv]# yum install net-snmp
Loaded plugins: dellsysid, fastestmirror
Loading mirror speeds from cached hostfile
-----
-----
Dependencies Resolved

========================================================================================================================================================================
Package    Arch         Version          Repository        Size
========================================================================================================================================================================
Installing:
net-snmp     x86_64   1:5.3.2.2-22.el5_10.1    updates  708 k
Installing for dependencies:
 lm_sensors     x86_64  2.10.7-9.el5       base     525 k
Updating for dependencies:
 net-snmp-libs  i386    1:5.3.2.2-22.el5_10.1    updates  1.3 M
 net-snmp-libs  x86_64  1:5.3.2.2-22.el5_10.1    updates  1.3 M

Transaction Summary
========================================================================================================================================================================
Install      2 Package(s)
Update       2 Package(s)
Remove       0 Package(s)

Total download size: 3.8 M
Is this ok [y/N]: y
Downloading Packages:
(1/4): lm_sensors-2.10.7-9.el5.x86_64.rpm        | 525 kB     00:01
(2/4): net-snmp-5.3.2.2-22.el5_10.1.x86_64.rpm   | 708 kB     00:02
(3/4): net-snmp-libs-5.3.2.2-22.el5_10.1.i386.rpm      | 1.3 MB     00:04
(4/4): net-snmp-libs-5.3.2.2-22.el5_10.1.x86_64.rpm    | 1.3 MB     00:03
------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Total   168 kB/s | 3.8 MB     00:23
Running rpm_check_debug
Running Transaction Test

Finished Transaction Test
Transaction Test Succeeded
Running Transaction
----
----
Installed:
  net-snmp.x86_64 1:5.3.2.2-22.el5_10.1

Dependency Installed:
  lm_sensors.x86_64 0:2.10.7-9.el5

Dependency Updated:
  net-snmp-libs.i386 1:5.3.2.2-22.el5_10.1 net-snmp-libs.x86_64 1:5.3.2.2-22.el5_10.1

Complete!
[root@umserv]#

2. Simple SNMP configuration:

mv /etc/snmp/snmpd.conf /etc/snmp/snmpd.conf.old

Add   below configuration to /etc/snmp/snmpd.conf

rocommunity  public  xxx.xxx.xxx.xxx
rocommunity  public   127.0.0.1
syslocation  "HYD, UM DataCenter"
syscontact  [email protected]

Replace xxx.xxx.xxx.xxx with the IP address of the server that you want to allow SNMP lookups from:
rocommunity public xxx.xxx.xxx.xxx

3. Start the SNMP service, and set it to auto-start on reboot:

/etc/init.d/snmpd start
chkconfig snmpd on

Note:If you have a firewall configured, ensure that you have UDP port 161 open to your SNMP lookup server.

4) Validation:

On your SNMP lookup server, you can do the following to perform a quick SNMP test to ensure that it’s working.

snmpwalk -v 2c -c public xxx.xxx.xxx.xxx or snmpwalk -v 1 -c public -O e 127.0.0.1
[root@umserv ~]# snmpwalk -v 1 -c public -O e 127.0.0.1
SNMPv2-MIB::sysDescr.0 = STRING: Linux umserv 2.6.18-92.1.17.el5 #1 SMP Mon Jul 14 06:07:13 IST 2014 i686
SNMPv2-MIB::sysObjectID.0 = OID: NET-SNMP-MIB::netSnmpAgentOIDs.10
DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (16748) 0:02:47.48
SNMPv2-MIB::sysContact.0 = STRING: [email protected]
SNMPv2-MIB::sysName.0 = STRING: umserv
SNMPv2-MIB::sysLocation.0 = STRING: "HYD, UM DataCenter"
SNMPv2-MIB::sysORLastChange.0 = Timeticks: (1) 0:00:00.01
...
...

Yes, it is working

Read More

AIX RC Scripts

We need some applications  should be stopped and started gracefully without manual intervention during the reboots . Order to serve this purpose , we use  rc scripts in all unix flavors including AIX  .

So, how do rc.scripts work:

1. Write a single script, put it into /etc/rc.d/init.d, make sure the script accepts a single parameter of start or stop and does the right thing.
2. In /etc/rc.d/rc2.d create a link (ln -s) to the script in init.d called Sxxname where xx is a number that dictates where in comparison to other scripts in the directory your script will execute (lower number first).
3. In /etc/rc.d/rc2.d create a link to the script in init.d called Kxxname where xx is a number which dictates when the script is run to stop your app in comparison to other scripts in the directory (lower number first).

Note: Its just convention to place scripts in /etc/rc.d/init.d and make  soft links  in /etc/rc.d/rc2.d. But its need not mandatory to keep  scripts in /etc/rc.d/init.d.

Example RC Script:

#!/usr/bin/ksh

ulimit -c 0

case "$1" in
start )
        ps -ef | grep -v grep | grep myengine > /dev/null
        ret=$?
        if [ $ret -gt 0 ]; then
                /var/myengine/bin/startup.sh
        fi
        ;;
stop )
        PID=$$
        for i in myengine-app1 myengine-app2 myengine-app3 myengine-app4; do
                ps -ef | grep $i | grep -v grep | awk '{print $2}' >> /tmp/myengine.$PID
        done
        while read line; do
                kill $line
        done < /tmp/myengine.$PID
        rm /tmp/myengine.$PID
        ;;
* )
        echo "Usage: $0 (start | stop)"
        exit 1
esac

Example Creating Symbolic Links

This is an example on creating symbolic links for automatic startup for tivoli. tivoli should start first (meaning a low Sxx) and stop last (meaning a high Kxx):

umadmin@umserve1:/etc/rc.d/rc2.d>sudo ln -s /etc/rc.d/init.d/rc.tivoli S20tivoli
umadmin@umserve1:/etc/rc.d/rc2.d>sudo ln -s /etc/rc.d/init.d/rc.tivoli K70tivoli

Read More