DT2283 Computer Systems Architecture and Administration

1
DT228-3Computer SystemsArchitecture and
Administration

• Mark Deegan
• Mark.Deegan_at_comp.dit.ie
• Room K115a

2
Aims of the CSAA Course

• To introduce the student to the general
  principles of computer systems administration.
  (S1)
• To equip the student with sufficient
  understanding of the architecture of computer
  systems to allow them to support systems
  administration activities. (S1)

3
Aims of the CSAA Course

• To provide a greater depth of knowledge of
  systems design and administration tasks. (S2)
• To give hands on experience of applying systems
  administration techniques. (S2)

4
Learning Outcomes

• On completion of the course the student will
• Have a broad understanding of Systems
  Architecture and Administration
• Be capable of analysing the architecture of a
  computer system and of understanding the effects
  of different architectural features on system
  administration.
• Be able to perform basic system administration
  tasks.
• Appreciate the protocols and policies of system
  administration

5
Examination

• Semester 1
• Written Examination 70
• Continuous Assessment(3 Assignments) 30
• Semester 2
• Written Examination 70
• Continuous Assessment(3 Assignments) 30

6
Reading List

• Essential Reading
• Running Linux
• Matt WELSH
• OReilly UK
• Background Reading
• Linux System Administration A Users Guide
• Marcel GAGNE
• Addison Wesley
• Linux System Administration Guide
• WIRZENIUS and OJA

7
Reading List

• Computer Organisation and Architecture
• William STALLINGS
• Prentice Hall
• Upgrading and Repairing PCs
• Scott MEULLER
• QUE
• Hardware Bible
• Winn L ROCHE
• QUE

8
CSAA Semester 1 Overview
9
CSAA Semester 1 Overview

• Essential Tasks of the System Administrator
• Booting and Shutting Down
• Controlling Processes
• File Systems
• UNIX User Management
• Windows User Management

10
CSAA Semester 1 Overview

• Logon Procedures
• Log Files
• Printing
• Network Configuration
• Performance Analysis
• UNIX Windows Co Operation

11
CSAA Semester 2 Overview
12
CSAA Semester 2 Overview

• Disk Management
• Backups
• Drivers and the Kernel
• Performance Analysis
• Planning and Configuring Network Services

13
CSAA Semester 1 Details
14
Essential Tasks of the System Administrator

• Adding Removing Users
• Adding and Removing Hardware
• Performing Backups
• Installing New Software
• Monitoring the System
• Managing System Resources
• CPU, memory, disk I/O, disk space
• Troubleshooting

15
Essential Tasks of the System Administrator

• Maintaining local documentation
• Security
• Contingency Planning
• Disaster Recovery
• Helping Users

16
Booting Shutting Down

• Bootstrapping
• Automatic Manual Booting
• Steps in the Boot Process
• Kernel Initialisation
• Hardware Configuration
• System Processes
• Scripts
• Startup Scripts

17
Booting Shutting Down

• Rebooting Shutting Down
• Turning Off the Power

18
Controlling Processes

• Components of a Process
• PID Process ID Number
• PPID Parent PID
• UID User ID
• EUID Effective User ID
• GID Group ID
• EGID Effective Group ID
• Niceness
• Control Terminal

19
Controlling Processes

• The Lifecycle of a Process
• Sending Signals to Processes
• Process States
• Nice and Renice
• Process Priority
• ps
• Monitoring processes
• top
• Runaway Processes

20
The File System

• Pathnames
• Mounting and Unmounting File Systems
• The Organisation of the File Tree
• File Types
• Regular Files
• Directory Files
• Character and Block Special Files

21
The File System

• UNIX Domain Sockets
• Named Pipes
• Symbolic Links
• File Attributes
• The setuid and setgid bits
• The Sticky bit
• Permission bits

22
The File System

• Viewing File Attributes
• Free BSD Bonus Flags
• Chmod Change Mode
• Chown Change Ownership
• Chgrp Change Group
• Umask Default Permissions
• Default Permissions

23
UNIX User Management

• The /etc/passwd file
• Login name
• Encrypted password
• UID Number
• Default GID Number
• GECOS Field
• Home Directory
• Login Shell

24
UNIX User Management

• Setting an Initial Password
• Creating users home directory
• Copying in the default startup files
• Setting the users mail home
• Editing the /etc/group file
• Setting Disk Quotas
• Verifying the New Login

25
UNIX User Management

• Removing Users
• Disabling Logins
• Vendor Supplied Account Management Utilities

26
Windows User Management

• LDAP Protocol
• Microsoft Active Directory
• Novell Active Directory
• Open LDAP
• User Account Creation
• User Management Tools
• Domain Structures

27
Logon Procedures

• UNIX Logon Procedures
• Character Mode
• Graphical More
• Dtlogin on Solaris and other UNIX variants
• Windows Logon Procedures
• Windows Telnet Services
• Windows Terminal Services

28
Log Files

• Logging Policies
• Disposal of Log Files
• Rotating Log Files
• Archiving Log Files
• Finding Log Files
• Condensing Log Files to Useful Information

29
Printing

• Types of Printers
• Serial Parallel Printers
• Network Printers
• Configuring the Printing Environment
• Controlling Spooling
• Controlling a Printer
• Adding a Printer
• Debugging Printer Problems

30
Network Configuration

• Configuring clients on an IP Network
• Static IP
• DHCP
• Bootp
• DNC Client configuration

31
Performance Analysis

• Optimising Performance
• Factors influencing performance
• System Performance Checkup
• Analysing CPU Performance
• How UNIX Manages Memory
• Analysing Memory Usage
• Analysing Disk I/O
• Procinfo

32
Performance Analysis

• Displaying Red Hat Performance Data
• Print random Free bsd statistics

33
Cooperating with Windows

• File and Print Sharing
• The Netwok File System
• CIFS The Common Internet File System
• Samba CIFS For UNIX
• Installing and Configuring SAMBA
• Debugging Samba
• Secure Terminal Emulation with SSH
• XWindows Emulators

34
Cooperating with Windows

• PC Mail Clients
• PC Backups
• Dual Booting
• Running Windows Applications under UNIX
• PC Hardware Tips

35
CSAA Semester 2 Details
36
Drivers and the Kernel

• Kernel Types
• Configuring a Kernel
• Build the Kernel
• Install the New Kernel
• Test The New Kernel
• Document the New Kernel
• Configuration File

37
Drivers and the Kernel

• Adding Device Drivers
• Device Numbers
• Device Files
• Naming Conventions for Devices
• Loadable KernelModules

38
Adding a Disk

• Disk Interfaces
• SCSI
• IDE
• SATA
• USB
• Disk Geometry

39
Adding a Disk

• Disk Installation Procedure
• Connecting the Disk
• Creating Device Entries
• Mounting a file system
• Formatting the disk
• Labelling and partitioning the disk
• Establishing logical volumes

40
Adding a Disk

• Software RAID
• Hardware RAIS
• Network Area Storage (NAS)
• Storage Area Networks (SANs)

41
Backups

• Labelling Tapes
• Selecting a backup Interval
• Comparison of file systems
• Making file systems smaller than the backup
  device
• Protection of Backups
• Checking Backup Tapes
• Developing a tape life cycle

42
Backups

• Designing data for backups
• Backup devices and media
• Setting up an incremental Backup Regime
• Dumping file systems
• Restoring individual files
• Restoring entire file systems
• Dumping and restoring for upgrades
• Commercial backup suites
• ArcServe, Veritas

43
Planning and Configuring Network Services

• Firewall
• SSHD
• BIND
• NIS
• DHCP
• OpenLDAP
• SaMBa

44
Lecture 1UNIX User Management

• Essentials tasks of the System Administrator
• Adding/Removing Users
• UNIX Traditionally, user account details are
  stored in the /etc/passwd file (/etc/group,
  /etc/shadow)
• This file has the format
• usernamepassworduidgidcommenthome dirshell
• one way to get a list of usernames on a system
• cut d -f1 /etc/passwd
• Do not edit this file directly, use
• useradd / userdel / groupadd / groupdel commands

45
useradd (linux)

• Create a new user or update default new user
  information
• useradd -c comment -d home_dir
• -e expire_date -f
  inactive_time
• -g initial_group -G
  group,...
• -m -k skeleton_dir -M -n
  -o -p passwd -r
• -s shell -u uid login
• useradd -D -g default_group -b default_home
• -e default_expire_date -f
  default_inactive
• -s default_shell
• The D option causes system defaults specified in
  /etc/login.defs to be used for this account
  creation

46
userdel

• Delete a user account and related files
• userdel -r login
• The userdel command modifies the system
  account files, deleting all entries that refer to
  login. The named user must exist. The options
  which apply to the userdel command are
• -r Files in the users home directory will
  be removed along with the home directory itself
  and the users mail spool. Files located
  in other file systems will have to be searched
  for and deleted manually.

47
groupadd (linux)

• Create a new group
• groupadd -g gid -o -r -f group
• Creates a new group account using the values
  specified on the command line and the default
  values from the system.
• The new group will be entered into the system
  files as needed.
• Options
• -g gid The numerical value of the groups ID.
  This value must be unique, unless the -o option
  is used. The value must be non-negative. The
  default is to use the smallest ID value greater
  than 500 and greater than every other group.
  Values between 0 and 499 are typically reserved
  for system accounts.

48
groupadd (linux)

• -r instructs groupadd to add a system
  account. The first available gid lower than 499
  will be automatically selected unless the -g
  option is also given on the command line.
• -f the force flag. This will cause groupadd
  to exit with an error when the group about to
  be added already exists on the system. If that is
  the case, the group wont be altered (or
  added again).
• Also modifies the way -g option works. When you
  request a gid that it is not unique and you dont
  specify the o option too, the group creation
  will fall back to the standard behavior (adding a
  group as if neither -g or o options were
  specified).

49
groupdel (linux)

• Delete a group
• groupdel group
• The groupdel command modifies the system
  account files, deleting all entries that refer to
  group. The named group must exist.
• You must manually check all filesystems to insure
  that no files remain with the named group as the
  file group ID.

50
groupdel (linux)

• Delete a group
• groupdel group
• The groupdel command modifies the system
  account files, deleting all entries that refer to
  group. The named group must exist.
• You must manually check all filesystems to insure
  that no files remain with the named group as the
  file group ID.

51
Managing Users

• Whether in the UNIX or Windows (or other)
  environment, it is best practice to manage users,
  where possible, on a per-group basis
• Wherever possible, settings and configurations
  should be managed on a per-group basis
• For this reason, the management of group settings
  is as important in any multi-user environment as
  is the management of individual settings

52
Managing Groups

• One approach to the management of groups is a
  hierarchical approach, whereby the user community
  is divided into a number of groups based on roles
  within the organisation
• In this way, an individual may belong to several
  groups, each of which may have a different
  combination of security settings etc.
• If we can be assured users have only setting
  which are assigned on a per-group basis, it
  becomes easier for us to ensure that all users
  with similar or identical roles are subject to
  the same rules

53
Lecture 2Windows User Management

• Brief History of Windows Server Platforms
• Server Message Block (SMB) Protocol
• Windows net commands

54
Managing Users and Groups in the Windows
Environment

• Since the arrival of Microsoft/IBMs Lan Server
  application in the late 1980s, Microsoft have
  provided a network server platform based on the
  SMB (Server Message Block) protocol
• The management tools and the features of these
  server platforms have changed somewhat over the
  years, but remain broadly the same

55
History

• Lan Manager
• Windows for WorkGroups
• Windows NT
• Windows Server 2000
• Windows Server 2003

56
Similarities between SMB based platforms

• Due to the similarities which exist between the
  various SMB based platforms, some knowledge of
  how to manage an SMB platform will serve will in
  many situations
• One of the most powerful SMB commands is the net
  command discussed later

57
Net command

• In the Microsoft Windows environment, the net
  command accepts a wide range of sub commands to
  achieve many varied tasks

58
Net sub-commands

• accounts, computer config, continue, file, group,
  help, helpmsg, localgroup, name, pause, print,
  send, session, share, start, statistics, stop,
  time, use, user, view
• Many of these, in turn, have further sub commands

59
Net accounts

• Updates the user accounts database and modifies
  password and logon requirements for all accounts.
  The Net Logon service must be running on the
  computer for which you want to change account
  parameters.
• net accounts /forcelogoffminutes no
  /minpwlenlength /maxpwagedays unlimited
  /minpwagedays /uniquepwnumber /domain
• net accounts /sync /domain

60
Net computer

• Adds or deletes computers from a domain database.
  This command is available only on computers
  running Windows 2000 Server.
• net computer \\computername /add /del

61
Net config

• Displays the configurable services that are
  running, or displays and changes settings for a
  service.
• net config service options

62
Net group

• Adds, displays, or modifies global groups on
  Windows 2000 Server domains. This command is
  available for use only on Windows 2000 Server
  Domain Controllers.
• net group groupname /comment"text" /domain
• net group groupname /add /comment"text"
  /delete /domain
• net group groupname username ... /add
  /delete /domain

63
Net localgroup

• Adds, displays, or modifies local groups.
• net localgroup groupname /comment"text"
  /domain
• net localgroup groupname /add /comment"text"
  /delete /domain
• net localgroup groupname name ... /add
  /delete /domain

64
Net name

• Adds or deletes a messaging name (sometimes
  called an alias), or displays the list of names
  the computer will accept messages for. The
  Messenger service must be running to use net
  name.
• net name name /add /delete

65
Net send

• Sends messages to other users, computers, or
  messaging names on the network. The Messenger
  service must be running to receive messages.
• net send name /domainname /users
  message

66
Net session

• Lists or disconnects the sessions between a local
  computer and the clients connected to it.
• net session \\computername /delete

67
Net share

• Creates, deletes, or displays shared resources.
• net share sharename
• net share sharenamedrivepath /usersnumber
  /unlimited /remark"text"
• net share sharename /usersnumber unlimited
  /remark"text"
• net share sharename drivepath /delete

68
Net start

• Starts a service, or displays a list of started
  services. Service names of two or more words,
  such as Net Logon or Computer Browser, must be
  enclosed in quotation marks (").
• net start service

69
Net statistics

• Displays the statistics log for the local
  Workstation or Server service, or the running
  services for which statistics are available.
• net statistics workstation server

70
Net stop

• Stops a Windows 2000 network service.
• net stop service

71
Net time

• Synchronizes the computer's clock with that of
  another computer or domain. Used without the /set
  option, displays the time for another computer or
  domain.
• net time \\computername /domaindomainname
  /rtsdomaindomainname /set
• net time \\computername /querysntp
  /setsntpntp server list

72
Net use

• Connects a computer to or disconnects a computer
  from a shared resource, or displays information
  about computer connections. The command also
  controls persistent net connections.
• net use devicename \\computername\sharename
  \volume password /userdomainname\use
  rname /delete /persistentyes no
• net use devicename /homepassword
  /deleteyes no
• net use /persistentyes no

73
Net user

• Net user
• Adds or modifies user accounts or displays user
  account information.
• net user username password options
  /domain
• net user username password /add options
  /domain
• net user username /delete /domain

74
Net view

• Displays a list of domains, a list of computers,
  or the resources being shared by the specified
  computer.
• net view \\computername /domaindomainname
• net view /networknw \\computername

75
Lecture 3Backup Restore

• Backup
• Restore
• tar, cpio, dd, dump/ufsdump
• Backup types
• Backup Plan
• Backup Scheduling

76
Backups

• The correct management of backup policy is
  possibly the most important role of a systems
  administrator
• It serves to provide
• Archives of data and state, roll-back, comparison
• Disaster recovery plan
• System replication mechanism
• Software deployment mechanism

77
Backup Restore

• Traditionally backup is to tape devices for
  reasons of capacity and reliability
• Backup may be to any other device depending on
  factors such as
• Capacity
• Speed
• Reliability
• Facility
• Cost
• Cost

78
tar

• Oldest UNIX backup method, runs on all UNIX
  variants
• Supports appending of files to the end of the
  tape
• Very slow, not really suitable for full system
  backups

79
tar

• User and group IDs are retained upon restore
• cd /export/home/michael
• tar cvf /dev/rmt/0m .
• tar xvf /dev/rmt/0m .

80
cpio

• Much faster than tar
• Good for replicating directory trees
• Supports incremental backup
• find . cpio o F archive
• cpio i F archive

81
dd

• Bit-for-bit copy
• Does not copy files and ownerships, just copies
  bits
• Cannot select files and directories from a dd
  tape
• Used for creating mirror images of disks at a low
  level

82
dump / ufsdump

• Used for creating full or incremental backups
• Can specify levels for dumping which specify the
  type of incremental backup
• Level 0 full backup
• Level 1 all files changed since last level 0
  dump
• Level 2 all files changed since last level 1
  dump
• Etc.

83
Backup Types

• Full
• Backup everything on the system
• Incremental
• Backup only those files that have changed since
  the last full backup
• Differential
• Backup only those files that have changed since
  the last backup

84
Grand Father - Father - Son Backup Scheme

• Perform an initial Full System Backup
• Perform daily backups, Monday, Tuesday,
  Wednesday, Thursdayreuse these tapes each week
• Perform weekly backups Friday1, Friday2,
  Friday3reuse these tapes each month
• Perform monthly backups Month1, Month2 etc.

85
Grand Father - Father - Son Backup Scheme

• Files can be restored to within 1 day, for files
  archived within the last week
• Files can be restored to within 1 week for files
  archived within last month
• Files can be restored to within 1 month for files
  archived within last year

86
Scheduling Backups

• Cron is the UNIX scheduling daemon
• The windows environment includes the Scheduler
  service which can be accessed using the At command

87
Cron (unix)

• Depending on the system configuration, users may
  schedule tasks to occur at specific times on a
  UNIX system
• The cron service searches the /var/spool/cron
  directory for cron files
• Cron files are named after usernames in
  /etc/passwd
• Files specify the time and the command to perform

88
Cron (unix)

• The cron service wakes up every minute and
  examines all the cron files
• Within each file it examines each line to
  ascertain if a command needs to be executed
• When the command is executed, any output is
  mailed to the user who owns the file or the
  username (if any) specified in the command

89
Cron (unix)

• Format of the crontab files
• minute (0-59)
• hour (0-23)
• monthday (1-31)
• month (1-12)
• weekday (0, Sunday 6, Saturday)
• username
• Command

90
Cron (unix)

• Example
• ls l /etc gt /home/mdeegan/etclist.txt
• 0 2 6 full-backup
• 15 12 1-5 incremental-backup
• 0 3 0 find / -name core atime 7 gt /corelist
• 0,20,40 0,6 w gtgt /home/mdeegan/wholog
• 0,10,20,30,40,50 1-5 w gtgt
  /home/mdeegan/wholog

91
Cron (unix)

• edited with crontab e
• displayed with crontab l
• deleted with crontab -r

92
At (unix)

• Used for single command execution at 900pm
  atgt find / -name core exec rm \

93
At (windows)

• at \\computername time /interactive
  /everydate,... /nextdate,... command
• at 1527 wholog.bat
• at 1700 /nextF collectCSAA.bat
• at 0300 /everyM,T,W,Th incBackup.bat
• at 0300 /everyF fullBackup.bat

94
Lecture 4Booting Linux

• Booting Linux
• LILO
• lilo.conf
• Loadlin.exe
• System Startup Shutdown

95
Booting LINUX

• Booting Linux requires that you install a program
  to load the kernel into memory on your computer
• Depending on your hardware, you may use one of
  several programs to do this
• LILO(intel), MILO(alpha), SILO(sparc)

96
Lilo

• You can use lilo to start Linux
• From the Master Boot Record (MBR) of your hard
  drive
• From the superblock of your linux root partition
  on a hard drive
• From a floppy disk

97
Drive Naming in Linux

• In Linux, as in other UNIX systems your hard
  drives are represented by files in the /dev
  directory
• If you have more than one IDE drive, the first is
  referred to as /dev/hda, the second as /dev/hdb
  etc.
• If you use SCSI drives, they will be named
  /dev/sda /dev/sdb etc.

98
Partition Naming in Linux

• Partitions within drives are referred to as the
  hard drive name plus a partition number /dev/hda1
  etc. or /dev/sda1 etc. for SCSI drives

99
Lilo.conf

• The operation of lilo is configured using the
  /etc/lilo.conf file
• Before configuring LILO you must know what
  partitions have what operating system on them
• For example If you have a dual-boot installation
  with 2 operating systems installed, /dev/hda1
  might be your Linux partition and /dev/hda2 might
  be your Windows 2000 partition

100
Lilo.conf

• You should know where you want to install LILO
• In most cases, you will want to put LILO on the
  MBR (Master Boot Record)

101
Lilo.conf

• In certain cases, you cannot install LILO on the
  MBR, you must install it in the Superblock of the
  root partition for your Linux installation
• LILO is normally installed after you have
  partitioned your hard drives and after you have
  installed either Linux or other operating systems

102
Editing Lilo.conf

• Editing lilo.conf is easy, but dangerous if you
  dont know what youre doing
• Always make a backup of lilo.conf (preferably to
  a bootable Linux floppy) before making any changes

103
Editing Lilo.conf

• You may want to edit lilo.conf for a number of
  reasons
• You are testing a new kernel and you want to be
  able to boot the same Linux partition with more
  than one kernel
• You want to add password protection to a
  partition
• You have a hardware setup that requires you to
  specify special options such as booting a remote
  file system
• Your kernel is called something other than
  /vmlinuz or is is a non-standard place such as
  /etc or /home

104
Sample lilo.conf

• Start LILO global section
• BOOT /dev/hda
• Prompt
• Vga normal
• Ramdisk 0
• End LILO global section
• image /vmlinuz
• root /dev/hda3
• label linux
• read-only
• other /dev/hda1
• label windows2000
• table /dev/hda

105
Lilo.conf parameters

• There are over 20 parameters that can be used in
  the lilo.conf file. Some of these are
• bootltboot_devicegt
• installltboot_sectorgt
• messageltmessage_filegt
• verboseltlevelgt
• backupltbackup_filegt
• prompt
• timeoutltsecsgt
• passwordltpasswordgt

106
/sbin/lilo

• After making any changes to lilo.conf you need to
  run /sbin/lilo
• This should also be run after installing a new
  kernel

107
Lilo boot prompt options

• The following options can be passed to LILO at
  the boot prompt to enable special features of
  your system or to pass options to the Linux
  kernel
• Normally, at the lilo prompt you will just type
  linux or dos
• Some other options that can be used are
• Rescue
• Single
• rootltdevicegt
• vgaltmodegt (can use the ask option)

108
Loadlin.exe

• Instead of using LILO, you can start Linux from
  DOS using loadlin.exe
• This is included on the installation CD-Rom in
  the Dosutils directory
• Loadlin.exe uses the DOS MBR to boot Linux
• Copy loadlin.exe to a DOS partition
• Put a copy of the Linux kernel in the DOS
  partition
• At the dos prompt type
• loadlin c\vmlinuz root/dev/hda1 rw

109
Diskless Linux workstation

• You may also choose to boot your PC as a diskless
  Linux workstation, provided the correct network
  infrastructure is in place

110
Lecture 5Linux StartUp and ShutDown

• Linux startup Shutdown
• The Boot Process
• The Initialisation Process
• Startup Scripts
• Run Levels
• Init, Shutdown, Halt, Reboot

111
The Boot Process

• Intel based PCs start by executing code contained
  in BIOS chips
• This code instructs the system to look for the
  first sector of the first cylinder of the boot
  drive and to try to load and execute code found
  there
• It is possible to set (in BIOS) the order in
  which the drives are searched (CD, Floppy, IDE1,
  IDE2 etc.)

112
The Boot Process

• This is also the case in many (but not all) other
  versions of hardware and of UNIX
• Linux shares many traits with System V and with
  BSD based UNIX variants
• Linux is closer to System V than to BSD in
  relation to the booting process
• Linux uses the init command and a directory
  structure based on run-levels to start running
  the system and loading processes

113
The Initialisation Process

• System States (or run levels) on early UNIX
  systems grew out of a need to separate how the
  system ran depending on the maintenance being
  performed
• Usually, this meant that to add new hardware (or
  software) a system reboot was necessary
• Today, with hot swappable devices, it is not
  strictly necessary to reboot systems in order to
  install new hardware or software

114
The initialisation Process

• The /etc/inittab file contains settings for the
  init process and also defines the run levels for
  a Linux system

115
/etc/inittab run levels

• Default runlevel. The runlevels used by RHS
  are
• 0 - halt (Do NOT set initdefault to this)
• 1 - Single user mode
• 2 - Multiuser, without NFS (The same as 3, if
  you do not have networking)
• 3 - Full multiuser mode
• 4 - unused
• 5 - X11
• 6 - reboot (Do NOT set initdefault to this)

116
/etc/inittab file (1/5)

• inittab This file describes how the INIT
  process should set up
• the system in a certain
  run-level.
• Author Miquel van Smoorenburg,
  ltmiquels_at_drinkel.nl.mugnet.orggt
• Modified for RHS Linux by Marc
  Ewing and Donnie Barnes

117
/etc/inittab file (2/5)

• Default runlevel. The runlevels used by RHS
  are
• 0 - halt (Do NOT set initdefault to this)
• 1 - Single user mode
• 2 - Multiuser, without NFS (The same as 3, if
  you do not have networking)
• 3 - Full multiuser mode
• 4 - unused
• 5 - X11
• 6 - reboot (Do NOT set initdefault to this)

118
/etc/inittab file (3/5)

• id3initdefault
• System initialization.
• sisysinit/etc/rc.d/rc.sysinit
• l00wait/etc/rc.d/rc 0
• l11wait/etc/rc.d/rc 1
• l22wait/etc/rc.d/rc 2
• l33wait/etc/rc.d/rc 3
• l44wait/etc/rc.d/rc 4
• l55wait/etc/rc.d/rc 5
• l66wait/etc/rc.d/rc 6

119
/etc/inittab file (4/5)

• Trap CTRL-ALT-DELETE
• cactrlaltdel/sbin/shutdown -t3 -r now
• When our UPS tells us power has failed, assume
  we have a few minutes
• of power left. Schedule a shutdown for 2
  minutes from now.
• This does, of course, assume you have powerd
  installed and your
• UPS connected and working correctly.
• pfpowerfail/sbin/shutdown -f -h 2 "Power
  Failure System Shutting Down"
• If power was restored before the shutdown
  kicked in, cancel it.
• pr12345powerokwait/sbin/shutdown -c "Power
  Restored Shutdown Cancelled"

120
/etc/inittab file (5/5)

• Run gettys in standard runlevels
• 12345respawn/sbin/mingetty tty1
• 22345respawn/sbin/mingetty tty2
• 32345respawn/sbin/mingetty tty3
• 42345respawn/sbin/mingetty tty4
• 52345respawn/sbin/mingetty tty5
• 62345respawn/sbin/mingetty tty6
• Run xdm in runlevel 5
• x5respawn/etc/X11/prefdm -nodaemon

121
init process

• init is sometimes referred to as the father of
  all processes
• init is responsible for creating processes from a
  script stored in /etc/inittab
• Linux init command is compatible with the System
  V init command
• init starts as the last step of the kernel
  booting
• init is the first command that initialises and
  configures the system for use

122
init process

• Init works by parsing the /etc/inittab file and
  by running scripts in the /etc/rc.d directories
  depending on the required run level
• Each individual script is designed to start or
  stop and individual service such as networking,
  mail, news, web, nfs, dhcp etc.

123
/etc/rc.d directory

• The /etc/rc.d directory contains the following
  files
• rc, rc.local, rc.sysinit
• And the following directories
• init.d
• rc0.d
• rc1.d
• rc2.d
• rc3.d
• rc4.d
• rc5.d
• rc6.d

124
Startup Scripts

• One of the most important scripts in /etc/inttab
  is rc.sysinit
• When init parses the inittab file, rc.sysinit is
  the first script found and executed

125
rc.sysinit script

• Sets some initial PATH variables
• Configures Networking
• Sets up Swapping for Virtual Memory
• Sets the system hostname
• Checks root file systems for possible repairs
• Checks root filesystem quotas
• Turns on user and group quotas for root file
  system

126
rc.sysinit script

• Remounts the root filesystem read/write
• Clears the mounted filesystems table /etc/mtab
• Enters the root filesystem into mtab
• Readies the system for loading modules
• Finds module dependencies
• Checks filesystems for possible repairs
• Mounts all other file systems

127
rc.sysinit script

• Cleans out /etc/mtab /etc/fastboot /etc/nologin
  etc.
• Deletes UUCP lock files
• Deletes stale subsystem files
• Deletes stale pid files
• Sets the system clock
• Turns on swapping
• Initialises the serial ports
• Loads Modules

128
rc.local script

• After the rc.sysinit script is run, init runs the
  rc.local script to carry out any site-specific
  operations that may be required upon system
  startup
• rc.local may examine the local system and create
  a message for the login prompt based on the
  hardware and software versions found
• If it does, it will put such details in the file
  /etc/issue which may be displayed on login

129
Run Level 0

• /etc/rc.d/rc0.d
• Starts the shutdown sequence
• Kills All Processes
• Turns off Virtual Memory File Swapping
• Unmounts swap and mounted file systems

130
Run Level 1

• /etc/rc,d/rc1.d
• Single user mode, or administrative state
• Used by sys admins while performing software
  maintenance
• Nobody else can log in
• Networking is turned off, but file systems are
  mounted

131
Run Level 2

• /etc/rc.d/rc2.d
• Multiuser state
• Networking is enabled, NFS is disabled

132
Run Level 3

• /etc/rc.d/rc3.d
• Usually the default run level specified as the
  first line in the /etc/inittab file
• Remote file sharing is enabled along with all
  other desired services

133
Run Level 4

• /etc/rc4.d
• Usually empty, not used
• If you want to define your own run level, this
  can be used to set up the appropriate links

134
Run Level 5

• /etc/rc.d/rc5.d
• Similar to the default run level, but with the
  named (internet DNS server)
• Usually the default run level for X11

135
Run Level 6

• /etc/rc5.d
• Reboot run level
• Contains links similar to level 0, but logic in
  the halt script in init.d determines whether the
  system is being shut down or rebooted

136
Run Levels

• Each run-level is defined by the services
  (daemons) that are normally running in that level
• The services to stop and start for each run level
  are defined in the rcX.d directories mentioned
  above
• These directories contain symbolic links to
  master scripts contained in the /etc/rc.d/init.d
  directory

137
Run Levels

• The symbolic links contained in the rcX.d
  directories must be named so as to start with
  capital K or capital S
• For example rc3.d contains the following
  symbolic links
• K05saslauthd K87portmap S10network S25netfs
  S80sendmail S95anacron
• Any other files or symbolic links contained in
  the rcX.d directories will be ignored by the rc
  script

138
Run Levels

• In addition to the inclusion of K or S at the
  start of the symbolic links, the file names also
  carry a number between 00 and 99
• The combination of K or S followed by a number
  means that when the files are listed
  alphabetically, they appear in a particular order
• This order determines the order in which the
  scripts are run by the rc script

139
Run Levels

• When a symbolic link beginning with K is
  encuntered, rc runs that script with the stop
  parameter
• When a symbolic link beginning with S is
  encountered, rc runs that script with the start
  parameter

140
/etc/rc/d/init.d directory

• This directory contains the master copies of all
  the scripts that may contain symbolic links in
  the rcX.d directories

141
Adding Services to Run Levels

• If we want a service to start in a given run
  level we must
• Create or copy the start/stop/status script to
  the /etc/rc.d/init.d directory
• Create a symbolic link in the required run level
  directory to start the script in the appropriate
  sequence
• ln -s /etc/rc.d/init.d/isdn /etc/rc.d/rc3.d/S20is
  dn

142
Init
143
Shutdown

• shutdown -krhfnc -t secs time warning
  message
• -k Dont really shut down, just warn
• -r reboot after shutdown
• -h halt after shutdown
• -f do a fast reboot
• -n do not go through init but go down real
  fast
• -c cancel a running shutdown
• -t ltsecgt delay between warning and kill signal
• Shutdown t60 r now

144
Halt

• Notifies the kernel of a shutdown or reboot

145
Reboot

• Symbolic link to halt

146
Lecture 6CSAA Lab 5

• 1 Directory Creation
• All work done as part of this lab should be
  stored in the following directory
• Windows U\MD-CSAA\LAB05
• or
• UNIX ltyour home directorygt/MD-CSAA/LAB05

147
Lab

• 2 Background
• This lab is intended to reinforce lectures on the
  topic of Linux system start-up.
• The lab helps to develop an understanding of the
  workings of the rc start-up script on a Linux (or
  other UNIX) system

148
Lab

• 3 Existing Files Directories
• On the Aisling server the directory
  /home/deegan-DT228-3/rc.d contains files and
  directories similar to those found in the
  /etc/rc.d directory
• init.d contains master copies of the scripts to
  be used as part of this lab

149
Lab

• The rc script  functions in a similar manner to
  the Linux rc script with the following exception
• Instead of the /etc/rc.d/rcltrun-levelgt.d
  directory it checks your own lab directory
  /home/student/DT228-3/ltyougt/MD-CSAA/LAB05/rc.d/rclt
  run-levelgt.d

150
Lab

• 4 Directories to Create
• In the LAB05 directory, you must create the rc.d
  directory
• In rc.d you must create run level directories,
  rc1.d, rc2.d and rc3.d
• In the rc3.d directory you must create symbolic
  links to effect the execution of the master
  scripts in the order specified in the rc3.d
  directory at /etc/rc.d/rc3.d
• To do this you must create correctly named
  symbolic links to the master files contained in
  /home/deegan-DT228-3/rc.d /init.d

151
Lab

• 5 Test
• You may test your work by running the rc script
  with the parameter indicating the required run
  level, for example
• /home/deegan-DT228-3/rc.d/rc 3

152
Lab

• 6 Completion Criteria
• When you have created the required symbolic links
  in the correct directories
• When you have successfully tested these links
  using the rc script
• You have successfully completed this lab

153
Lecture 7Performance Analysis

• Performance Analysis
• vmstat

154
Performance Analysis

• Performance analysis involves identifying various
  system bottlenecks
• This incolves a number of steps
• We must ask a number of questions
• Is there a performance Problem?
• Is the problem CPU or I/O related?

155
CPU Related?

• What is the current load on the CPU?
• What is the average load on the CPU?

156
I/O Related?

• Is it normal disk I/O?
• Would more/faster disks help?
• Is it paging I/O?
• Would more physical memory help?

157
Related to a Particular User or Program?

• Identify the user / program
• Identify what they are doing to cause the problem
• Revise their operating procedures
• Consider removing them from the system

158
Determining CPU Usage

• Determining the CPU usage is the first thing we
  should do
• There are a number of tools to do this
• vmstat gives several pieces of useful information
  including CPU usage
• vmstat interval count
• Interval is the number of seconds between reports
  and count is the number of reports to generate

159
vmstat 2 10

• mdeegan_at_aisling mdeegan vmstat 2 10
• procs memory swap
  io system cpu
• r b w swpd free buff cache si so
  bi bo in cs us sy id
• 1 0 0 5484 27240 136584 198840 0 1
  5 8 8 8 4 7 4
• 0 0 0 5484 27240 136584 198840 0 0
  0 96 155 100 0 0 100
• 0 0 0 5484 27232 136584 198844 0 0
  0 0 159 112 2 0 98
• 0 0 0 5484 27216 136584 198844 0 0
  0 0 130 51 0 2 98
• 0 0 0 5484 27216 136588 198848 0 0
  0 86 157 63 0 0 100
• 0 0 0 5484 27216 136588 198848 0 0
  0 0 139 46 0 0 100
• 0 0 0 5484 27224 136588 198836 0 0
  0 30 153 47 0 0 100
• 0 0 0 5484 27712 136588 198824 0 0
  0 8 166 107 1 0 99
• 0 0 0 5484 26876 136588 198828 0 0
  0 0 139 92 6 2 91
• 0 0 0 5484 26876 136592 198824 0 0
  0 144 137 69 0 0 100

160
vmstat

• The first line gives the average values since the
  system was booted and should be ignored
• To determine the CPU usage, we are interested in
  the last three columns, us, sy, id

161
Analysing vmstat output(Process States)

• There are three states in which a process may be
  at any point in time
• Runtime, uninterrupted sleep, swapped out
• Process Statistics
• r Number of processes waiting for runtime
• b Number of processes in uninterrupted sleep
• w Number of processes swapped out, but otherwise
  able to run

162
Analysing vmstat output(Process States)

• The number of processes waiting for runtime is a
  good indicator of system health
• The more processes waiting, the slower will be
  the system

163
Analysing vmstat output (Memory)

• Memory Statistics
• swapd Amount of virtual memory used (KB)
• free Amount of idle memory (KB)
• buff Ammount of memory used in buffers
• cacheamount of memory left in cache

164
Analysing vmstat output (Swap)

• Swap Statistics
• si Amount of memory swapped in from disk (KB/s)
• so Amount of memory swapped out to disk (KB/s)
• Swap statistics are arguably the most important
  statistic to monitor, and of these, the so field
• This field indicates the pages that have been
  swapped out, even if done before vmstat was
  started

165
Analysing vmstat output (I/O)

• I/O Statistics
• bi Blocks received from a block device
  (blocks/sec)
• bo Blocks sent to a block device (blocks/sec)
• If there are a large number of block transfers,
  the problem with your system may lie here
• A single reading, however is not indicative of
  the system as a whole, simply a snapshot
• All Linux blocks are 1KB except for CDRom blocks
  (2KB)

166
Analysing vmstat output (System)

• System Statistics
• in The number of interrupts per second,
  including the system clock
• cs The number of context switches per second

167
Analysing vmstat output (I/O)

• I/O Statistics
• bi Blocks received from a block device
  (blocks/sec)
• bo Blocks sent to a block device (blocks/sec)
• cs The number of context switches per second
• If there are a large number of block transfers,
  the problem with your system may lie here
• A single reading, however is not indicative of
  the system as a whole, simply a snapshot

168
Analysing vmstat output (CPU)

• CPU Statistics
• us Percentage of CPU cycles spent on performing
  user tasks
• sy Percentage of CPU cycles spend on system
  tasks. These include writing I/O, performing
  general O/S functions etc.
• id Percentage of CPU cycles not used (idle)

169
Analysing vmstat output (CPU)

• Just because CPU time is high or idle time is low
  does not indicate a system problem
• It may simply indicate that a number of batch
  jobs are scheduled to run at the same time and
  might benefit from being rearranged
• In order to establish if there is a genuine
  problem it is necessary to monitor the system
  over an extended period
• If average CPU remain high, there is a problem

170
Lecture 8Performance Analysis

• Performance Analysis
• top

171
top

• top is another tool for identifying problems with
  a LINUX system
• Displays the top CPU processes
• Displays a listing of the most CPU intensive
  tasks on the system
• Can provide an interactive interface for
  manipulating the processes
• Default is to update every 5 seconds

172
top

• top operates by examining files in the /proc
  pseudo file system
• This pseudo file system is used as an interface
  to kernel data structures
• man proc

173
top

• mdeegan_at_aisling mdeegan top
• 171441 up 47 days, 227, 8 users, load
  average 0.06, 0.03, 0.07
• 61 processes 59 sleeping, 2 running, 0 zombie, 0
  stopped
• CPU states 0.0 user 0.2 system 0.0 nice
  0.0 iowait 99.8 idle
• Mem 513316k av, 200052k used, 313264k free,
  0k shrd, 44976k buff
• 57692k actv, 11208k in_d,
  1024k in_c
• Swap 1052248k av, 9096k used, 1043152k free
  34656k cached
• PID USER PRI NI SIZE RSS SHARE STAT CPU
  MEM TIME CPU COMMAND
• 1 root 15 0 108 76 56 S 0.0
  0.0 015 0 init
• 2 root 15 0 0 0 0 SW 0.0
  0.0 000 0 keventd
• 3 root 15 0 0 0 0 SW 0.0
  0.0 001 0 kapmd
• 4 root 34 19 0 0 0 SWN 0.0
  0.0 000 0 ksoftirqd_CPU0
• 9 root 15 0 0 0 0 SW 0.0
  0.0 000 0 bdflush
• 226 root 15 0 0 0 0 SW 0.0
  0.0 000 0 kjournald
• 586 root 15 0 200 160 116 S 0.0
  0.0 008 0 syslogd
• 590 root 15 0 180 168 120 S 0.0
  0.0 003 0 klogd
• 666 root 15 0 480 348 232 S 0.0
  0.0 109 0 sshd
• 719 root 15 0 52 4 0 S 0.0
  0.0 000 0 gpm

174
Analysing top output

• Up The time the system has been up and the three
  load averages
• Average number of processes ready to run in the
  last 1,5 and 15 minutes
• Same as the output of uptime
• Processes The total number of processes running
  at the time of the last update
• Broken down into running, sleeping, stopped and
  zombied

175
Analysing top output

• CPU States The percentage of CPU time in user
  mode, system mode, niced tasks (negative nice
  tasks) and idle
• Time spent in niced tasks will also be counted
  system and user time, so the total will be more
  than 100
• Mem Statistics on memory usage, including total
  available memory, free memory, used memory,
  shared memory, memory used for buffers

176
Analysing top output

• Swap Statistics on swap space including total
  swap space and used swap space
• This and the Mem section together are the same as
  the output of free
• PID The process ID of each task
• USER The username pf the tasks owner
• PRI The priority of the task
• NI The nice value of the task. Negative values
  are lower priority

177
Analysing top output

• SIZE The size of the tasks code plus data stack
  space, in kilobytes
• RSS The total amount of physical memory used by
  the task in kilobytes
• SHARE The amount of shared memory used by the
  task
• STATE The state of the task, S sleeping, D
  uninterrupted sleep, R running, Z zombies, T
  stopped or traced

178
Analysing top output

• CPU The tasks share of the CPU since the last
  screen update as a a percentage of total CPU time
• MEM The tasks percentage of physical memory
• Time Total CPU time used by process since it
  started
• COMMAND The tasks command name

179
Using top to control processes

• In addition to command-line options for
  controlling the appearance of top (not covered
  here) there are a number of commands that can be
  issued to top while running
• Space immediately updates the display
• L Erases and redraws the screen
• k kill a process You will be prompted for the
  pid and a signal to send to the process (normally
  15)

180
Using top to control processes

• i ignore zombie processes
• n change the number of processes to view
• r renice a process
• P sort tasks by CPU usage
• M sort tasks by Memory usage

181
Renice

• The renice command is used to alter the priority
  of running processes
• The default nice value is 0
• The range in Linux is -20 to 20
• The lower the value the faster the process runs
• Can examine the nice value of a process using ps
  l

182
Renice

• The owner of and root can change the nice value
  of aprocess using renice
• Changes apply to all child processes
• renice priority -p pid ... -g pgrp ...
  -u user ...
• mdeegan_at_aisling mdeegan ps -l
• F S UID PID PPID C PRI NI ADDR SZ WCHAN
  TTY TIME CMD
• 0 S 1634 24496 24495 0 75 0 - 1091 wait4
  pts/1 000000 bash
• 0 R 1634 26361 24496 0 75 0 - 778 -
  pts/1 000000 ps
• mdeegan_at_aisling mdeegan renice 5 24496
• 24496 old priority 0, new priority 5
• mdeegan_at_aisling mdeegan ps -l
  
  
• F S UID PID PPID C PRI NI ADDR SZ WCHAN
  TTY TIME CMD
• 0 S 1634 24496 24495 0 80 5 - 1091 wait4
  pts/1 000000 bash
• 0 R 1634 26363 24496 0 80 5 - 777 -
  pts/1 000000 ps

183
Renice

• Once a nice value has been increased, only the
  root user can reduce it again, not even to the
  default value
• mdeegan_at_aisling mdeegan renice 19 24496
• 24496 old priority 5, new priority 19
• mdeegan_at_aisling mdeegan ps -l
  
  
• F S UID PID PPID C PRI NI ADDR SZ WCHAN
  TTY TIME CMD
• 0 S 1634 24496 24495 0 94 19 - 1091 wait4
  pts/1 000000 bash
• 0 R 1634 26390 24496 0 94 19 - 778 -
  pts/1 000000 ps
• mdeegan_at_aisling mdeegan renice 1 24496
  
  
• renice 24496 setpriority Permission denied

184
Lecture 9File Systems

• Guest Lecture by Emmet Caulfield
• File Systems

185
Lecture 10RPM Package Management

• Guest Lecture by Emmet Caulfield
• Package Management

186
Lecture 11UML On McCool

• User Mode Linux (UML) On McCool

187
User Mode Linux (UML)

• User-Mode Linux is a safe, secure way of running
  Linux versions and Linux processes.
• Run buggy software, experiment with new Linux
  kernels or distributions, and poke around in the
  internals of Linux, all without risking your main
  Linux setup.
• http//user-mode-linux.sourceforge.net/

188
User Mode Linux (UML)

• UML is a useful teaching tool, especially in
  courses where students need a dedicated machine
  in order to get the most out of it
• UML is being used to teach OS development,
  network administration, and more general system
  administration.
• These are all cases where having a virtual
  machine to practice on is far more convenient for
  everyone than using physical boxes.
• http//user-mode-linux.sourceforge.net/

189
UML at the School of Computing

• At present (01/12/2004) The School of Computing
  provides one system specificall for hosting UML
  sessions
• McCool.student.comp.dit.ie

190
McCool

• For ease and speed of initial set up the
  /etc/passwd and /etc/group files were copied from
  aisling.student.comp.dit.ie
• Home directories for students were also copied
  from aisling
• The net effect is that the initial passwords and
  initial directory contents of student accounts on
  McCool were the identical