Sunday, March 28, 2010

Change UUID of virtual drive

I made a copy of virtual drive (by copy and pasting .vdi file). Then, I tried to mount the .vdi file in the virtual machine. It came up with some error: A hard disk with UUID {bla..bla..bla......} is already registered. So, I was unable to mount the drive.

Now, what is the solution?
Ans: Change the UUID of the virtual/image drive.

How can I change the UUID of the virutal drive?
Ans: You have to run the VBoxManage command

[Before running command, make sure that environment variable is set to the location of the virtual box commands. To set environment variable: Go to "My Computer" --> Right click --> Properties --> Advanced --> Environment Variables --> Add the location of Virtual Box to the PATH variable; In my case it is : PATH .......;C:\Program Files\Sun\VirtualBox ]

C:\ > VBoxManage internalcommands setvdiuuid disk2.vdi

[disk2.vdi is the name of my copied .vdi file]

Now, try to mount the disk2.vdi in Virtual Machine.
Have a good one!


On my second day, I found issue with SLES (Suse Linux). It keeps on complaining "VB is waiting for /dev/disks/by-id/scsi-xxxxx-part2 to appear".
After some research on blogs, found that GRUB in SLES looks for UUID of disk to boot.

I set the copied/cloned harddisk as second harddisk (primary slave) for the original SLES VM. Then I startet the original SLES VM.
In the running SLES VM I made a new directory and mounted it to the root directory of the cloned harddisk.
Then I startet (as root) the Yast partition manager and noticed the UUID of the cloned harddisk. (It is made from the UUID shown with vboxmanage list hdds, but it is not the same).
Then I edited (as root) the /boot/grub/menu.lst and the /etc/fstab : I changed the UUID everywhere the old one was (do not use capital letters instead of small letters).
Now shutdown (after unmounting the cloned root directory)
Then I disabled the cloned harddisk as second harddisk in the original VM and set it as first harddisk in the cloned VM.
Starting the cloned VM now works.


For Network cards to work you have to do
  1. rename the ifcfg-eth-id- to ifcfg-eth0 (in /etc/sysconfig/network)
  2. remove the rules file (/etc/udev/rules.d/30-
  3. edit /etc/sysconfig/network/config and change FORCE_PERSISTENT_NAMES to "no".

Saturday, March 27, 2010

Install VMware-tools

It might be challenging to install VMware-tools if we don't know what exactly we are doing.
You need to read all the instructions that is provided by VMware workstation or Virtual Center or ESX server.
For now, I am going to give the instruction for Linux systems.
After installing Linux as Virtual Machine on the top of VMware workstation/ESX Server or VSphere (whatever), you will see that VMware-tools is not installed.

Steps to be followed:

1) Choose "Install VMware tools" from VM menu

2) You will recognize some change on "cd-rom" icon located at Left corner of VM. It just like someone has inserted the CD of VMware tools in your VM.

3) Now you need to mount the CD-ROM so that you can access the contents.
#mount /dev/cdrom /mnt/cdrom
If you are lucky, above command will work without any issue. If not you might have to specify some mount options like
# mount -o ro /dev/cdrom /mnt/cdrom
(which mean mount as read-only option)
Some systems requires you to specify the file-systems type. FYI, CD-ROM uses iso9660 file-system
# mount -t iso9660 -o ro /dev/cdrom /mnt/cdrom

4)After this, go to /mnt/cdrom directory. You will find VMware-toools over there. You can install from RPM or TAR file. It all depends on you, however I recommend tar file.

# cd /mnt/cdrom
# rpm -ivh [VMware-tools......rpm]

or
# cd /mnt/cdrom
#tar -xvzf [VMware-tools.......tar] /root/vmware
(this will un-archive the "tar" file under /root/vmware directory
# cd /root/vmware
#./vmware-install.pl

Tuesday, March 9, 2010

Some slick Unix/Linux commands... much more powerful and useful

KILL

usage:

Kill process by Process ID

  • kill -9 [PID] --> Forcefully Kill
  • kill -15 [PID] --> Gracefully Kill
  • kill %[JobID] --> JobID can be checked by command #jobs

Kill process by Process Name

  • killall [proces name]

PID/Process name can be found using commands like

  • ps -aux
  • ps -elf
  • netstat -anp
  • lsof -w -n -i [tcp/udp:port number]

System Tuning Commands:

#mpstat -A ALL --> CPUs status

#top --> Press 'I' to toggle between Irix Mode (Irix Mode display resource usage mulitplied by total number of CPUs) ; Press "1" to see all the CPUs

# sar

Networking Commands

#netstat -anp --> Where 'p' flag shows which program is making connection

[ Hey, now I know which program is making a connection, can I use kill command to kill by process name or PID? Answer: Ofcourse, that's the reason I am writing this article ]

Now you know how to terminate the particular remote connection in the server. There is also another technique using #lsof.

# lsof -w -n -i tcp:22

This command will display all the SSH (tcp:22) connections. So, you can see who has intruded in your system using SSH. This will also give process ID (PID) or process name, which you can terminate using KILL command.

[ use man page for each command in details]

#ifstat

#ethtool [ethernet interface name] --> this command can used to see if the interface is physically down or not

e.g # ethtool eth1

#netstat -i

#netstat -s


To get the BIOS information

#dmidecode

To list the hardware

#hwinfo

or

#lshw

[Note: lshw is not the linux project. If you want to use lshw, you have to download and install the RPM for that command http://ezix.org/project/wiki/HardwareLiSter ]


Bash Script to collect system information:

This is one of the simple script file to pull basic and informative information about your server:


#********************

#ServerReport.sh

# Created by DShah

# Please use it at your own risk

#********************

#!/bin/bash
miniDivider(){

echo "*******************************************************************";
}

serverInfo() {

uname -a
cat /etc/*release

}

diskInfo() {

# fdisk -l
df -h
iostat

}

memoryInfo() {

free -m
vmstat

}

cpuInfo() {

mpstat
mpstat -P ALL

}

overallInfo() {

top -b -n2


}

networkInfo() {
netstat -s
miniDivider

sleep 2
netstat -s
}

dividerLine(){

echo "########################################################################";

echo "########################################################################";

}

fileName="serverReporting.txt"

rm $fileName
touch $fileName

echo "ServerInfo Reporting"
serverInfo >> $fileName
dividerLine >> $fileName

echo "DiskInfo Reporting"
diskInfo >> $fileName
dividerLine >> $fileName

echo "MemoryInfo Reporting"
memoryInfo >> $fileName
dividerLine >> $fileName

echo "CPUInfo Reporting"
cpuInfo >> $fileName
dividerLine >> $fileName

echo "Network Info Reporting"
networkInfo >> $fileName
dividerLine >> $fileName

echo "Top command reporting";
overallInfo >> $fileName