AIX Disk Help

AIX Disk Overview

You measure a Real Computer by weight and by current it draws. A Real Computer is measured in tons and amps. A Real Computer will weigh at least a ton and suck down enough juice to baroque a rhino. We're talking python cables, raised floors, haylon dumps, and explosive bolt firing power supplies. Real Computers are feared by mere mortals and even the experienced wizard approaches with respect.
Real Computers have disk drives the size of washing machines. When a Real Computer is plugged in the lights dim. Real Computers don't get turned off, just rebooted every few years. The docs for Real Computers come on palet jacks and have been known to kill whole forests.-- Jeff Whyche (posting on alt.folklore.computers)

OVERVIEW OF DISKS

At the very basic level, all information is held on hard-disks within or attached to the UNIX box. As far as the user is concerned, they just have files, and do not care which disk the file ends up on. Also, we do not want to place a restriction on file size - if someone wishes to have a file 8GB big then this should be possible, even if we only have 2GB disks (say). This functionality is achieved by having physical disks (the actual disks within the drive), volume groups (groups of physical disks) and logical volumes (areas of a fixed size).
At the very bottom level, we have the physical disks themselves. To get a list of the disks in the current machine type 'lspv', which will produce a list like this:

$ lspv
hdisk0	0009388a0025f169	rootvg
hdisk1	000b388d97a3987e	imclvg
hdisk2	00200964953fbe0a	extvg2
hdisk3	00200964953ede8c	extvg1
hdisk4	00200964953ef112	extvg1
hdisk5	00200964953fd16d	extvg2
hdisk6	00200964953fd934	extvg2
hdisk7	00200964953fe0f3	extvg2
hdisk8	00200964953fe8bd	extvg2
hdisk9	00200964953ff081	extvg2

The first column shows the name of the disk, the second column the location and the third column will show which volume group the hard disk is in. To find out more about a particular disk, use the command 'lspv <disk name>' for example:
$ lspv hdisk0

PHYSICAL VOLUME	hdisk0	VOLUME GROUP: rootvg
PV IDENTIFIER	0009388a0025f169	VG IDENTIFIER 0009388a0a93fd00
PV STATE	active
STALE PARTITIONS	0	ALLOCATABLE: yes
PP SIZE	16 megabyte(s)	LOGICAL VOLUMES: 8
TOTAL PPs	542 (8672 megabytes)	VG DESCRIPTORS: 2
FREE PPs	439 (7024 megabytes)
USED PPs	103 (1648 megabytes)
FREE DISTRIBUTION	108..63..52..108..108
USED DISTRIBUTION	01..45..56..00..01

Descriptions of the fields:

PHYSICAL VOLUME	<name of the disk>	VOLUME GROUP: <VG the disk belongs to>
PV IDENTIFIER	?	VG IDENTIFIER: ?
PV STATE	<shows whether disk is in use>
STALE	?	ALLOCATABLE: ?
PP SIZE	Physical partition size (very similar to Cluster sizes on DOS)
LOGICAL VOLUMES	<How many logical volumes are on each disk>
TOTAL PPs	show size of the disk	VG DESCRIPTORS: ?
FREE PPs	show amount of disk space left
USED PPs	show amount of disk space used

DISTRIBUTIONS: relate to the edge, middle, and centre of the disk. Every disk, because it is circular is divided very roughly into 3 rings. Consider the fact that the read-write head of the disk has, on average, less time to travel to the middle of the disk than to the edge or the centre. Therefore, the operating system will try and locate everything along the middle of the disk to ensure best possible speed. The distributions field shows what files are located where (and I apologise for the very poor diagram). On AIX distributions given in lspv, the disk is split into 5 rings and the usage of each is indicated in the 01...45...56...00....01 listing, starting from the edge and working in.
Disk Diagram

Finally, you may see exactly which 'logical volumes'/'file systems' are located on the physical disk by using 'lspv -l <disk name> e.g.
$ lspv -l hdisk0 hdisk0:

LVNAME	LPs	PPs	DISTRIBUTION	MOUNT POINT
hd5	1	1	01..00..00..00..00	N/A
hd6	32	32	00..32..00..00..00	N/A
hd2	31	31	00..13..17..00..01	/usr
hd8	1	1	00..00..01..00..00	N/A
hd4	2	2	00..00..02..00..00	/
hd9var	10	10	00..00..10..00..00	/var
hd3	25	25	00..00..25..00..00	/tmp
hd1	1	1	00..00..01..00..00	/home

The above shows that which logical volumes are located on hdisk0, and also their distributions with the disk, and how much space is allocated to each filesystem on this particular disk (seen by viewing # of PPs).

To view all the volume groups that exist, use the command:
$ lsvg
rootvg
extvg1
imclvg
extvg2

To view specific characteristics of a volume group, use the command
lsvg -l <volume group name> . e.g.
$ lsvg -l rootvg
rootvg:

LVNAME	TYPE	LPs	PPs	PVs	LV STATE	MOUNT POINT
hd6	paging	32	32	1	open/syncd	N/A
hd5	boot	1	1	1	closed/syncd	N/A
hd8	jfslog	1	1	1	open/syncd	N/A
hd4	jfs	2	2	1	open/syncd	/
hd2	jfs	31	31	1	open/syncd	/usr
hd9var	jfs	1	10	1	open/syncd	/var
hd3	jfs	25	25	1	open/syncd	/tmp
hd1	jfs	1	1	1	open/syncd	/home

This list shows all the logical volumes within the volume group. LV NAME gives the name referred to by the operating system of the logical volume. TYPE can be paging (i.e. Swap space, used by UNIX as an overflow from memory), boot (contains the boot block, and is not accessible normally), jfs (stands for journalled file system, and is the place generally where files are kept) and jfslog (shrug. Don't know what this is). LPs - stands for Logical Partitions and is almost always the same as PPs unless the volume is mirrored on a seperate disk. e.g.
$ lsvg -l extvg1
extvg1:

LV NAME	TYPE	LPs	PPs	PVs	LV STATE	MOUNT POINT
loglv01	jfslog	1	2	2	open/syncd	N/A
lv01	jfs	368	736	2	open/syncd	/usr/live

$ This is a mirrored drive, so the number of PPs (Physical Partitions) is double the number of LPs (Logical Partitions). PVs stands for 'Physical Volumes' and is the number of disks within this particular volume group. LV STATE?. MOUNT POINT is the directory and sub-directories (though not all of them if they are a different logical volume) that the user sees and places files in.

So, if I create a file '/usr/live/x.txt', it goes into the logical volume lv01, and lv01 is in the volume group extvg1 which contains two disks, and as this file system is mirrored, it ends up on both disks (hdisk3 and hdisk4).

Finally to check the status of your logical volumes, just type the command 'df -k' to see the usage of logical volumes:
$ df -k

Filesystem	1024-blocks	Free	%Used	Iused	%Iused	Mounted on
/dev/hd4	32768	20468	38%	1144	7%	/
/dev/hd2	507904	105736	80%	29708	24%	/usr
/dev/hd9var	163840	153436	7%	486	2%	/var
/dev/hd3	409600	216836	48%	3140	4%	/tmp
/dev/hd1	16384	7368	56%	2859	70%	/home
/dev/c2000	2048000	640436	69%	9140	2%	/usr/c2000
/dev/lv01	3014656	578564	81%	6123	1%	/usr/live
/dev/data	1540096	823104	47%	876	1%	/usr/live/data
/dev/arch01	1638400	251320	85%	787	1%	/usr/arch01
/dev/arch02	1638400	996852	40%	785	1%	/usr/arch02
/dev/arch03	1638400	103672	94%	791	1%	/usr/arch03
/dev/arch04	1638400	374068	78%	820	1%	/usr/arch04
/dev/arch05	1638400	334680	80%	832	1%	/usr/arch05
/dev/arch06	1638400	315140	81%	828	1%	/usr/arch06
/dev/arch07	1638400	164240	90%	850	1%	/usr/arch07
/dev/arch08	1638400	371612	78%	852	1%	/usr/arch08
/dev/arch09	1744896	240084	87%	873	1%	/usr/arch09
/dev/arch10	1638400	145508	92%	809	1%	/usr/arch10
/dev/arch11	1638400	108084	94%	784	1%	/usr/arch11
/dev/arch12	1638400	229056	87%	812	1%	/usr/arch12
/dev/usr2	1343488	256024	81%	6278	2%	/usr2
/dev/t2000	2007040	557744	73%	7205	2%	/usr2/t2000
/dev/arc	819200	211696	75%	1609	1%	/usr2/arc
/dev/arch1996	73728	5132	94%	304	2%	/usr/arch1996
/dev/arch1997	155648	56316	64%	368	1%	/usr/arch1997
/dev/progs	368640	25784	94%	2311	3%	/usr/live/progs
/dev/demo	2048000	788820	62%	3843	1%	/usr/demo
/dev/clones	2949120	1678260	44%	4247	1%	/usr/clones

The filesystem is the name of the logical volume, the total size is in the field '1024-blocks'. Most fields here are self explanatory.

Paging space can be looked at using 'lsps':
$ lsps -a

Page Space	Physical Volume	Volume Group	Size	%Used	Active	Auto	Type
hd6	hdisk0	rootvg	512MB	32	yes	yes	lv

This shows all the paging spaces on the disks. Again, paging spaces are simply overflows from memory, and is also called (normally by windows users) swap space or virtual memory.

Please use SMIT to change anything to do with disks

When changing any volume groups it is imperative that you immediately create a system boot tape (mksysb). To create this, use smit mksysb -> Backup DEVICE or FILE = /dev/rmt0, Create MAP files: yes, EXCLUDE files?: no, Make BOOTABLE backup? yes, EXPAND /tmp if needed? yes

It is also wise to include a listing of the current volume groups, and their placings on various disks by using the above commands. If you do not make a boot tape after changing volume groups, it is unlikely that you will be able to recover the system if it crashes.

To increase the size of a volume group:
smit -> System Storage Manager (Physical & Logical Storage) -> Logical Volume Manager -> Logical Volumes -> Set Characteristics of a logical volume -> Change characteristics of a logical volume -> enter logical volume name -> Increase the value of 'MAXIMUM NUMBER OF LOGICAL PARTITIONS'
F3 -> Increase the size of a Logical Volume -> enter logical volume name -> Place a value in 'Number of ADDITIONAL logical partitions'

----------- End of DiskHelp Page ------------

28th September 2001