PCI RS/6000 Start up Process Overview

POST

Power On

Locate OS Bootstrap Image

Find AIX Boot Image

Load AIX Boot Image RAMDISK created

Configuration Manager Phase 1

INITNormal IPL

AIX Startup on itanium- based SystemIBM POWER processor Intel Itanium – based

processor

ROS

SMS / bootlist

BIOS / EFL

Setup EFI Boot Manager

•ROS / SOFTROS •EFI system partition boot.efi

•BLV •volume partition BLV

Hardware Initialization

Low level configuration

Load boot Image

AIX initialization

alog

User Applications Install ProcessBoot Process NIM

Alog program

/var/adm/ras/bootlog

/var/adm/ras/Bosmenus.log

/var/adm/ras/nosinst.data

.

.

/var/adm/ras/erriog

Use alog command to

view logs

To view the boot log:

#alog –o –t boot

System Resource Controller

Subsystem group spooler

Subsystem qdaemon writesrv ipd

Subserver qdaemon writesrv Ipd

•Provides a single interface to control subsystems

•Controls individual or groups of subsystems

System Resource Controller Syntax

# Lssrc –g spooler

Subsystem Group PID Status

Qdaemon Spooler 8022 active

Writesrv Spooler 9558 active

Lpd Spooler inoperative

# startsrc –s lpd

0513-059 The lpd Subsystem has been started. Subsystem PID is 12472.

# refresh –s lpd

0513-095 The request for subsystem refresh was completed successfully

# stopsrc –s lpd

0513 -044 The lpd Subsystem was requested to stop.

List SRC Status

Start a Subsystem

Refresh a Subsystem

Stop a Sybsystem

Device States

Predefined Database

Undefined Supported Device

Customized Database

Defined Not Usable

Available Ready for Use

MkdevOrcfgmgr

rmdev -dl

rmdev -dl

rmdev -l

Self-Configuring Devices

1.Who are you

2. Answer -CD – ROM-04 – C0-00 – 3,0

Device Driver

cfgmgr

ODM

Cd0

Defined

04-c0-00-3,0

4a) Load device driver4b) Make /dev/cd0 entry

Kernel /unix

Device Driver

22

ODM

Cd0

Available

04-c0-00-3,0

3.cd0 defined

5.Device available

# ls –l /dev/cd0Br—r—r– root System 22, 2 /dev/cd0

AB – CD – EF – G , H

AB – CD Identifies the bus and the adapter locationSame as with non SCSI devices

EF For a single SCSI bus -00For a dual SCSI bus Internal bus - 00 External bus – 00

G,H G – SCSI address ( SCSI ) of the deviceH = Logical unit number of the devices

Location Code Format for SCSI Devices

Location Code Example for SCSI Device

SCSI Devices (Disk, Tape, CD-ROM)

System Unit

SCSI Adapter

04-01

0

74

6

SCSI Bus

SCSI ID

04-01-00-4,0

04-01-00-6,0

Traditional UNIX Disk Storage

Partition 1

Partition 2

Free space

Partition 3

Free space

PROBLEMS:

•Fixed partitions

•Expanding size of the partition

•Limitation on size of a file system and a file

•Contguous data requirement

•Time and effort required in planning ahead

Logical volumes solve non – contiguous problems

Logical volumes can span disks

Dynamically increase logical volume size

Logical volumes can be mirrored

Hard disk easily added to a system

Logical volumes can be relocated

Volume group and logical volume statistics can be collected

These task can be performed dynamically!

Benefits of the LVM

Physical Storage

Volume Group A

Volume Group B

PV1

PV2 PV3 PV4 PV5

PP1

PP2

PP3PP4

PP5

PP.

PP.

PPn

Physical Partitions 4MB is Default size 1016 max per PV

Physical Volume /dev/hdiskn

Volume Groups

PV1

PV2

PV4

PV3

datavgrootvg

New PVs:

•Add to existing VGs

•Create new VG

Why create new volume groups?

•Separate user data from operating system files

•Disaster recovery

•Data portability

•Data integrity and security

Volume Group Descriptor Area

VGDA

VGDAVGDA

VGDA VGDA VGDA VGDA

VGDA VGDA

Three disk or more VGTwo disk VG

One disk VG

Max size

Normal VG-32 PVs

Big Vg-128 PVs

Number of disks: Max.number of partitions/disk

1 32512

2 16257

4 8128

8 4046

16 2032

32 1016

Big Volume Groups (mkvg- B)

Number of disks: Max.number of partitions/disk

1 130048

2 65024

4 32512

8 16256

16 8128

32 4064

64 2032

128 1016

Volume Group LimitsNormal Volume Groups (mkvg)

Mkvg -t

Uses of Logical Volumes

A logical volume may contains one of the following , and only one at a time:

Journaled or Enhanced journaled file system ( for example : / dev hd4

Paging space ( dev/hd6 )

Journal log ( /dev/hd8 )

Boot Logical Volume ( /dev/hd5 )

Nothing ( raw device )

What Is File System ?

A File System is Method of storing data Hierarchy of directories

Four types supported : Journaled File system ( jfs2 ) Enchanced Journaled File System ( jfs2 ) CD – ROM File System ( cdrfs ) Network File System ( nfs )

Different file systems are connected together via directories to form the veiw of files users see

Why Have File System ?

Can strategically place it on disk for improved performance Some tasks are performed more efficiently on a file system than on each directory within the file system, for example , back up, move, secure an entire file system Can limit disk usage of users by file system ( quotas ) Maintain integrity of the entire file system structure, for example ,if one file is corrupted ,the others are not affectedSpecial security situations Organize data and programs into groups for ease of the file management and better performance

home sbin lib lpp usr bin dev tftpboot var mnt etc tmp

Standard File Systems in AIX

hd1 hd2 hd9var hd3

Bin lib lpp share sbin Spool adm tmp

hd4

(root)

Listing Logical Volume Information

List all Logical Volumes by Volume Group:

# lsvg –l rootvg

LVNAME Type Lps PPs Pvs Lv STATE Mount Point

Hd6 paging 64 64 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 156 156 1 open/syncd /Usr

Hd9var jfs 1 1 1 open/syncd /var

Hd3 jfs 3 3 1 open/syncd /tmp

Hd1 jfs 1 1 1 open/syncd /home

1v00 jfs2 2 2 1 open/syncd /home/john

Volume Groups

Volume Group

PV PV

Physical Volume (PV)

Volume Group (VG)

Hard disk

Collection of related disks (PVs)

1st copy of data1 LV

2nd copy of data1 LV

1st copy of data2 LV

2nd copy of data2 LV

Hot Spare

Synchronization

Hot Spare

First Copy

Second Copy

Third Copy

Mirroring

Hdisk()

PP1

PP2

Hdisk1

PP1

PP2

Hdisk2

PP1

PP2

LP1

LP2

Lv00

Read Request

Copy 1Copy 2 Copy 3

Forced Synhronization

Synchronize the Read Partitions

Mirror Write Consistency

E M C M E

Logical Volume PoliciesIntra-physical volume allocation policy:

Ineer Edge

Ineer Middle

Center

Outer Middle (Middle)

Center Edge (Edge)

Inter-physical volume allocation policy:

•Maximum number of PVs to use

•Range of PVs to use

I/O

Busy Disk

Less Busy Disk

Move Physical Partitions

# migratelp hd3/1 hdisk 1/109

Migrating Physical Partitions

Structure of a Journaled File System

Data Blocks

Indirect Blocks

inodes

Superblock

• Superblock

1. File System size and identification

2. Free list, fragment size, nbpi

• Inodes

1. File size, ownership, permissions, times

2. Pointers to data blocks

• Blocks1. Data blocks –contain data2. Indirect blocks –contain pointers to data

blocks.

Permission no.of links type of file user ID group ID file size address of blocks time modified time accessed time changed access control info. Reserved other

Inode for file 1

Structure of an Inode

•Contents of an inode

File System Fragmentation

No Fragmentation

File size = 2000 bytes

Fragmentation Enabled

File size = 2000 bytes

Fragment size = 1024 bytes

4096 bytes

2000 bytes 2000 bytes

This free space cannot be used by another file

1024 1024 1024 1024

These free fragments can be used by other files

2000 bytes

Variable Number of Inodes With the default nbpi=4096 on inode will be created for every 4096 bytes of file system.

File System

File System

1

2

3

4

5

6

7

8

1

2

Using the value nbpi=1024 an inode will be created for every 1024 bytes of file system.

4096 4096 4096

4096 4096 4096

128 bytes

128 bytes

Allocation Group Size

Groupings of related inodes and disk blocks.

Groupings of related inodes and disk blocks.

16 MB

16 MB

16 MB

64 MB

64 MB

64 MB

inodes

inodesDisk blocks

Disk blocks

16 MB

64 MB

agsize

agsize

JFS vs.JFS2 File Systems

JFS JPS2

Maximum File Size Architectural /Tested

64 Gigabytes /64 Gigabytes 4 Petabytes / 1 Terabyte

Maximum File Size Architectural /Tested

1 Terabyte / 1 Terabyte 4 Petabytes / 1 Terabyte

Inode Size 128 Byters 512 Bytes

Number of inodes Fixed. Set at creation Dynamic

Fragments/ Block Size 512/4096 512/4096

Online defragmentation Yes Yes

Avaiable on itaum No Yes

Avaiable on POWER Yes, default Yes

Large File Enabled File Systems

1

2

3

4

.

1023

1024

1025

1026

1027

.

1055

1056

1057

1058

1059

.

File = 132 MD

(1024 * 4 KB blocks)+ (1024 * 128 KB blocks)= 132MB

4MB + 128MB = 132MB

File 132 MB

32 Blocks 128 KB

32 Blocks 128 KB

128 KB

1 Block 4 MB

Fsync()

1) Inode changes to log

2) COMMIT to log

3) Update inode

4) Sync log

Write data

inodes

Data block JFSLOG

3

1

4

2

No journaling of data blocks –only journals inode information (and indirect block information.)

Journal Log

Space Management

File system expand upon notice, not automaticallyTo keep from running in to problems :Monitor file system growth Determine cause Control growing filesManage file system space usageControl user disk usage Defragment file system

Control Growing Files

/var/adm/wtmp/var/spool/*/*$HOME/smit.log$HOME/smit/script$HOME/websm.log$HOME/websm.script/etc/security/failedlogin/var/adm/sulog

Skulker

The skulker command cleans up file system by removing unwanted or obsolete filesCandidate files include ( can use file aging as criteria ):

Those in /temp directory A.out file Core files Ed.hup files

Skuler is normally invokes daily by the cron command as part of the root’s crontab fileModify the skulker shell script to suit local needs for the removal of files

What is Paging Space?

RAM – 256 MD

RAM Usage

Operating System Database TCP/IP 8MB FREE

Operating System Database TCP/IP Application

4KB 4KB

Paging Space

Current applications Total =248 MB

RAM Usage

•Only one paging space per disk

•Use disks with the least activity

•Paging spaces roughly the same size

•Do not extend paging space to multiple PV’s

•Use multiple disk controllers

hd6 paging00Paging01

Paging Space Placement

AIX Product Offerings

AIX

Expansion Pack

Bonus Pack

LPPs

AIX Documentation

AIX Toolbox for Linux

LOO:

bos

Package: bos.INed

Package: bos.adt

Fileset:

Bos.INed

Fileset:

Bos.adt.lib

Fileset:

Bos.adt.prof

Collection of Packages Complete Product

Collection of Filesets

Smallest Unit Specific Function

Packaging Definitions

LPP Package Fileset Suffix

Bos.terminfo.print.data

Message Convention:

LPP.msg[.lang].package.fileset

Fileset Naming

5 . 1 . 0 . 0

Version Release Modification Fix

MIGRATION Smit update_all

Software Updates

Applied:

Fileset 4.3.3.0 (Old Version)

Applied fileset 5.1.0.0

Committed Fileset 5.1.0.0

Commit or Reject

Fileset 5.1.0.0 (New Version)

Fileset 4.3.3.0 (Old Version)

Install

Install

Committed

4.3.3.0

5.1.0.0

Software States

Why Backup?•Data is very important:

Expensive to recreate

Can it be recreated?

•Disaster recovery:

Hardware failure

Damage due to installation/repair

Accidental deletion

•Transfer of data between systems Reorganizing file systems

•Defragmentation to improve performance

•System image for installation

•Checkpoint (befor/after upgrade)

•Long term archive

System

Records image backup of the operating system

Full Preserves all user data and configuration files

Incremental Records changes since previous backups Must be used carefully Very quick

Types of BackupThree types of backup:

rootvg

Full backup

System backup

Incremental backup

Incremental backup

Backup Strategy

AIX Print Subsystem : Advantages

Powerful and flexible printer drivers System management tools Limits fields and option validation Easy printer customization Single step print device and queue creation Support for dial-in administration Customizable spooling subsystem

System V Print Subsystem: Advantages

Compattability of interface programs

Avilabilty of interface programs

Security

Support for forms

Standard PostScript filters

Long term strategy

File 1

File 2

File 3

File 4

File 1

File 2

File 3

File 4

/dev/Ip0

/dev/Ip1

Queue2

Queue1

:

Concepts of Queues

Spool

Directory

Virtual Printer

Defination

Ip Ipr qprt

enq

Queue

qdaemon

Backend

( piobe )

/ev/Ip0d

# qprt -Pps [ -c ] file Print Request

monitors

starts

submits file to prienter

Print Request

Copy of file ( if requested )

Printer Data Flow

/etc/qconfig queue configuration files

/var/spool/* spooling directories

/var/spool/lpd/qdir/* queue requests

/var/spool/qdaemon/* temporary enqueued files

/var/spool/ipd/stat/* line printer status information

/var/spool/lpd/pio/@local Virtual printer directories

System Files Associated with Printing

State DescriptionDEV_BUSY Printer is busy servicing other print requests

DEV_WAIT QUEUE IS WAITING FOR THE PRINTER

DOWN QUEUE IS DOWN AND NO JOBS WILL BE SERVICED FROM THIS QUEUE UNTIL IS BROUGHT UP

OPR_WAIT THE QUEUE IS WAITING FOR OPERATOR INTERVENTION

QUEUED JOB IS QUEUED AND WAITING

READY Everything is ready to receive a print request

RUNNING Print file is printing

UNKNOWN Problem with the queue – need to investiage further to determine cause

Queue Status