Split and Joining a Volume Group in AIX

Splitting refers to the process of a breaking a mirrored Volume Group, which can be imported and used in a different AIX Machine or a LPAR. Applicaton Portability with less downtime are executed using the Splitting and Join Operation.

Scenario:

Consider an application configured on a volume group called "Application" which is made up of hdisk0 and hdisk1. And the application is mirrored.

Note: Only a mirrored volume group can be splitted.

Steps to Split:

1) # lsvg -o

2) # lsvg -l application /* Check the VG is mirrored */

3) # lsvg -p application  /* PV's used to make the Volume Group*/

4) # splitvg -y <new vgname> -c 2 application /* Splitting application VG to a new VG, if no new name is givethen the VG is created with the default name vg00 */

5) New VG name is given as say "backupappln"

6) The VG "backupappln" is created in the PV hdisk. Where hdisk1 was a part of mirrored VG application.

7) Verify the new VG Created.

8) # lsvg - l backupappln

9) # lsvg -l application

10)  # lsvg -p application

The output will still show that the VG application is made up of hdisk0 and hdisk1 but the state of hdisk1would be snapshotlv.


11) # lsvg -p backupappln

The output will show only the hdisk1

12) Now we export the VG backupappln to be used in a different machine. Before exporting the VG backupappln make sure that there is no Open Logical Volume and Varyoff the VG.
    
     # varyoffvg backupappln
 
13) # exportvg backupappln

14) # rmdev -Rdl hdisk1 /* The disk hdisk1 has been removed from the AIX Machine */

15) # lsdev -l hdisk1 /* Ensure the hdisk1 is not shown */

16) Connect the hdisk1 to the required machine.

17) # cfgmgr /* Look for the newly connected device */

18) # lsdev -l hdisk1 /* Check the status of hdisk */

19) # importvg -y VG Name PV Name

VG Name : backupappln and PV Name : hdisk1

20) # lsvg -o /* Output will show the VG backupappln */

21) # lsvg -l backappln /* List the LV's in the VG */

22) By default all the LV's  and mountpoints are prefixed with "fs".

23) # chlv -n newname oldname /* To change the name of the LV's

24) # chfs -m newmount point filesystem /* To change the mount point */

25) # mount -a /* To mount the file system *.

VIOS Operation-3

Management of Virtual/Shared Devices:

To create a virtual target device that maps lv20 as a virtual disk for a client partition hosted by the vhost0 virtual server adapter :

# mkvdev -vdev lv20 -vadapter vhost0

To create a virtual target device that maps hdisk5 as a virtual disk for a client partition hosted by the vhost1 virtual server adapter :

# mkvdev -vdev hdisk5 -vadapter vhost1

To create a SEA that maps the physical Ethernet adapter ent4 as a virtual adapter for the clien tpartitions served by the virtual Ethernet adapters ent6 and ent7, using ent6 as the default adapter and 8 as the default id :

# mkvdev -sea ent4 -vadapter ent6,ent7 -default ent6 -defaultid 8

Assign IP Address:

# mktcpip -hostname <Hostname> -interface ent# -inetaddr <IP> -netmask -gateway <IP> -start

To remove a virtual target device vtscsi9 along with its backing lv:

# rmdev -vdev vtscsi9

To remove all virtual target device associated with backing device lv001 :

# rmdev -vdev lv001

To remove a specific virtual target device :

# rmdev -vtd vtscsi0

VIOS Operation-2

Device Configurations:


1) # cfgdev /*To configures the devices in the VIO server*/


2) # cfgdev -dev  /*To configures the devices attached to scsi0 adapter*/


3) # chdev -dev <****> -attr <attribute=value> -perm /*To change the attribute of a device permanently*/


4) # chpath -dev hdisk1 -pdev scsi0 -op disable /*To disable the path between hdisk1 and scsi0*/


5) # rmdev -dev /dev/cd0 /*To unconfigure the cd-rom device*/


6) # rmdev -recursive -dev scsi /*To unconfigure scsi adapter and its child devices*/


7) # rmdev -pdev pci0  /* To unconfigures the children of pci1 but not the PCI bus pci1 */
 

VIOS Operation

Commands to Backup VIOS:

1) # backupios -cd /dev/cd0 -cdformat /*Backup VIOS on CD*/

2) # backupios -cd /dev/dvd -udf   /*Backup VIOS on DVD */

3) # backupios -tape /dev/rmt0 /*Backup VIOS on Tape*/

4) # backupios -file /tmp/filename /*Backup VIOS on a file*/

5) # backupios -tape /dev/rmt0 -verify /*Verify the Backup*/

HMC Operations

This posts covers the steps to be followed in taking backup of Manager Partition Profile data.

The Operations Performed are:

Backup * Restore * Initialize * Delete

1) Taking Backup of Managed Partition Profile

Open HMC->On the Workpane->Select System Management->Servers->Select the desired the Managed System.

2) Once the Managed System is selected->Check the Check Box->Click on the >> icon that appears on the Managed System.

3) Right click on the >> icon->Select "Manage Partition Data"->Select "Backup/Restore/Delete/Initialize".

System Dump

System Dump Device is used for Crash Recovery.

By default  /dev/hd6 is assigned for Dump.

Steps to create a dump device:

1) Estimate the amount of size required to create a dump device:

# sysdumpdev -e

The output is shown in bytes.

2) Check the free PP's available in rootvg.

# lsvg rootvg

Determine about PP's required to create dump device.

3) Check the PV's in the rootvg

# lsvg -p rootvg

Select the PV which is having more free PP's to create dump device.

4) Create a dump device.

# mklv -y dump_device -t dump rootvg PP's PV

5) Check the dump device

# lsvg -l rootvg

6) Initialzie the Dump

# sysdumpdev -K

7) Assign the dump device as a primary dump device

# sysdumpdev -Pp /dev/dump_device

8) verify the dump device

# sysdumpdev -l

House Keeping in HMC

This section shows how to handle filesystem "/var" utilization in an HMC. Since the /var filesystem holds all the log files, performing periodic house keeping on /var would keep it in a consistent state.

Login into HMC-CLI.

1) # monhmc -r {disk|proc|swap|mem} -n #

-r :  Resource
-n : Interval to repeat the status of the resources.

2) # chhmcfs -o f { -d days | -h hours| -s size} -f filesystem

-o : Operation to be performed, "f" refers to free the filesytem.

# chhmcfs -o f -d 3

The above command cleans up the log files from all the filesystem except the last 3 days -d 3.

# chhmc -o f /var

The above command cleans up all the log files in tehe /var filesystem.

3) # monhmc -r disk

Verify the filesystem size after performing the operation.

MultiBos - Operating System Backup

Multibos is one of the utilties in AIX to take a backup of the Operating System. Multibos is capable of creating multiple instance of the Operating System. Unlike cloning, it takes backup of the Operating System in the rootvg itself. All the filesystem are prefixed with "bos_"

Note: Login as a root user.

1) Ensure you have enough PP's in your rootvg and no stale PP's in the rootvg.

# lsvg rootvg

2) Initiate the multibos.

# multibos -sXp

-s :  Created an instance Standby BOS.
-X : Increase the size of the filesystem dynamically.
-p : Preview mode. Once preview is fine execute the same command without -p.

3) It would take 15 mins to complete.

4) Check the multibos standby BOS created by,

# lsvg -l rootvg

5) To mount all the standby BOS use the command,

# multibos -Xm

6) Operation on the multibos BOS can be done through multibos shell.

# multibos -S

7) Remove the standby BOS through,

# multibos -RXp

-p for preview operation. Once done with preview -p can be removed and executed.

RootVG Backup - alt_disk_copy method

Hi Friends, in this post I would like to discuss few tools used to take a backup of the Operating System Volume Group, when we perform some administration tasks like Technology Upgrade and Maintenance Level Upgrade.

Tool1:

# alt_disk_copy : Its a harware migration tool to take a backup of the Opeartion System. This method is alos called as "Cloning of OS". This method facilitates taking backup of the Operating System on a different Physical Volume ( Which is not assigned to any Volume Group). The rootvg created by this method has a prefix "altinst_" and the same for all the rootvg fileystems.
# lspv   /* Select any one unasssigned Physical Volume */

If you would like to exclude any file system and the filesystem entry has to be made in /etc/exclude.rootvg

# vi /etc/exclude.rootvg
^./ <filesystem> and save the file.

# alt_disk_copy -e -d hdisk# 

The process would take atleast 20 minutes to complete. Once its completed you can view the alt_inst_rootvg by executing,

# lsvg

# lspv

# lsvg -l alt_inst_rootvg

In order to perform any actions on the alt_inst_rootvg, we need to use the command alt_rootv_op

# alt_rootvg_op -W -d hdisk#  /* To wake up the alt_inst_rootvg */

# alt_rootvg_op -S -t hdisk#   /* To put the alt_inst_rootvg in sleep mode */

# alt_roortvg_op -X hdisk#  /* To remove the alt_inst_rootvg*/

# alt_rootvg_op -v <new name> -d hdisk# /* To rename the alt_inst_rootvg*/

Above alt_inst_rootvg could be updated in the bootlist using the bootlist command.

NTP Configuration- AIX Client and Server

AIX Server Configuration:

NTP : Network Time Protocol. Used for time synchronization to perform and support real-time applications.

1) vi /etc/ntp/.conf
# comment the Broadcast line

2) Include the entry,

server 127.127.X.Y, where Y refers to the statum value for the purpose of synchronization. Statum value should be 0 to 3. Save the file.

3) # startsrc -s xntpd

4) # lssrc -ls xntpd

AIX Client Configuration:

1) Ensure there is a proper communication between the client and server.

2) Edit the configuration file for NTP.

# vi /etc/ntp.conf

3) Include the line

server <IP Address of the NTP Server or Name of the NTP Server>

4) Save the file.

5) # startsrc -s xntpd

NTP Syncronization process will take ~10 mins.

6) # lssrc -ls xntpd, if the statum value is "16". It shows no syncronization between the NTP Client and Server.

7) In that case manually configure the NTP,

# ntpdate <IP Address of the NTP Server or Name of the NTP Server>

8) Verify the same.

# ntpq -a or ntpq -peers

Interview Questions Discussed

1) What is the command to view the active vg? 

# varyonvg -o

2) How do u configure the Disk? 


# cfgmgr


3) Tell me Steps to configure the VG? 


# cfgmgr
# lspv 
# mkvg -y vgname -s PP Size PV's
# lsvg -o


4) How do u add a new disk in VG? 


# cfgmgr
# extendvg vgname newdis(PV)
# lsvg -p vgname /* To verify the new disk successfully added to the VG */


5) What are the Attibutes of LVM? 


Attributes include Physical Volume, Physical Partition, Logical Volume, Logical Partition, Volume Group, File system.


6) How do u find the fix is installed? 

# instfix -icqv  <FIX NO>

7) How to u extend the FS? 

# chfs -a size=<+desired size> <filesystem>

8) Attributes of FS? 

Attributes include File System Type, Mount Point, Accounting and Logical Volume.

9) List all the LV in system? 

# lslv 

10) How do u find the PP size of the VG? 

# lsvg vgname

11) How do u create the pp size of 32 MB in VG? 

# mkvg -s 32 hdisk's

11) What is the Limitation of VG? 

Big VG, Normal VG and Scalable VG.

12) How do u disable the paging space? 

# chps -a n paging

13) What is LPAR?

LPAR is called Logical Parition. An object of Virtualization, which is an AIX/Linux/VIOS server which resources are logically shared and built.

LVM Commands

# lspv - List available Physical Volumes

# lspv -l PV - List the logical volume configured on the Physical Volume.

# lspv -M PV - View the mirroring information of the Physical Volume.

# lspv -p PV - View the physical partitions information on the Physical Volume.

# lquerpv -M PV - View the Logical track group of a Physical Volume.

# mkvg -y <VG Name> -s <Physical Partition Size in the power 2> <Physical Volumes> - To create a Volume Group with a specific name of desired PP size.

# mkvg -y <VG Name> <Physical Volumes> - To create a Volume Group with the specific name of default PP size (default 4 MB).

# mkvg -s <Physical Partition Size in the power 2> <Physical Volumes> - To create a Volume Group with the default name (default vg00) of desired PP Size.

# mkvg <Physical Volumes> - To create a Volume Group with default name of default size.

# lsvg - To view all the configured Volume Groups.

# lsvg -o - To view all the active Volume Groups.

# lsvg -l VG Name - To view the logical volumes of that Volume Group.

# lsvg -p VG Name - To view the physical volumes of that Volume Group.

# lsvg -M VG Mame - To check the Volume Group is mirrored or not.

# lsvgfs VG Name - To view the filesystems on the Volume Group.

Linux IP Address Bonding

Bonding refers to the process to binding an IP Address interface to one or more interface for redundancy purpose.

Here I have posted how to creating a bonded interface by binding eth0 and eth1 together.

Steps to create IP Address Bonding in Linux :

1) # cd /etc/sysconfig/network-scripts/

2) # touch ifcfg-bond0          /* Creating a IP address configuration file for Bonding inteface */

3) # vi ifcfg-bond0 and add the following information:

DEVICE = bond0
IPADDR = IP address of your interface
NETMASK = Corresponding netmask
GATEWAY = Gateway address
BOOTPROTO = none
ONBOOT = Yes

If you find difficult in typing this information, these informations could be copies from one of the configuration files like "ifcfg-eth#" and it can be edited.

4) Similarly edit, # vi ifcfg-eth0 and ifcfg-eth1

Ensure you  have only the below informations on both the files.

DEVICE = eth#
BOOTPROTO = yes
ONBOOT = yes
MASTER = bond0
SLAVE = yes

This ensures that the bond0 is used as a primary interface and eth0 and 1 act as a secondary interface.

5) Edit /etc/modprobe.conf

alias bond0 bonding
alias bond0 mode=balance_alb miimon=100

6) # modprobe bonding

7) # service network restart

8) # chkconfig network on

9) Restart the machine.

10) Finally verify the bonded interface in  # ifconfig -a

11) Can also be verified in  /proc/net/bonding/bond0.

Good to Know about AIX Commands-13/03/2011

1. # bootinfo -o hdisk # - To find the Device Address.
2. # svmon -P -t 1 - To find the process consuming more memory.
3. # export HISTTIMEFORMAT="%d%m%yyyy-%H%M%S" - To view the history with time stamp.
4. # export HISTSIZE=500 - Sets the size of the history file to 500.
5. # mkdir <filename> and # cd $! - This opens the file which you created.
6. # errlogger "Test Message" - To check the Error logger Demon is working fine.
7. # find / -xdev ls | sort  +6rn | head - To find the file/directory consuming more space.
8. # topas -P - To view only the process related information.
9. # lsrsrc "IBM Management Server" - To check whether your machine is managed by HMC.
10. # lvlstmajor <VG Name> - To list the major number of a Volume Group.

What is AIX ?

AIX is an open operating system from IBM that is based on a version of UNIX. AIX/ESA was designed for IBM's System/390 or large server hardware platform. AIX/6000 is an operating system that runs on IBM's workstation platform, the RISC 
System/6000.

AIX

This is Rajasekar.M. This blog is created to share vital information about IBM AIX and Linux among the working professional and the Beginners. The aim of creating this blog is to share knowledge on UNIX. Looking forward for your support.